Professional Documents
Culture Documents
Atlas Medical Ultrasonografie
Atlas Medical Ultrasonografie
June 2021
ULTRASONOGRAPHY
A N I N T E R N A T I O N A L J O U R N A L O F C L I N I C A L I M A G I N G
Taking ultrasound
to new heights
ACUSON Sequoia
ultrasound system
siemens-healthineers.com/ultrasound
Contents
Editorial
To screen or not to screen for NAFLD?
I. Sporea .................................................................................................................................................................................... 133
Original papers
Transfer learning with pre-trained deep convolutional neural networks for the automatic assessment of liver
steatosis in ultrasound images
E.C. Constantinescu, A-L. Udriștoiu, Ș.C. Udriștoiu, A.V. Iacob, L.G. Gruionu, G. Gruionu, L. Săndulescu, A. Săftoiu ....... 135
Contrast-enhanced ultrasound for the assessment of focal nodular hyperplasia – results of a multicentre study
R. Şirli, I. Sporea, A. Popescu, M. Dănilă, D.L. Săndulescu, A. Săftoiu, T. Moga, Z. Spârchez, C. Cijevschi, C. Mihai,
S. Ioanițescu, D. Nedelcu, N. Iacob, G. Miclăuș, C. Brisc, R. Badea ....................................................................................... 140
Evaluation of bowel preparation before colonoscopy by ultrasonographic monitoring of colonic fecal retention:
a case series
M. Matsumoto, M. Fujioka, T. Okada, Y. Naka, A. Amemiya, E. Matsushima, N. Tamai, Y. Miura, G. Nakagami,
H. Sanada .................................................................................................................................................................................. 147
The value of abdominal ultrasonography compared to colonoscopy and faecal calprotectin in following up
paediatric patients with ulcerative colitis
I. Fodor, O. Serban, D.E. Serban, D. Farcau, O. Fufezan, C. Asavoaie, S.C. Man, D.L. Dumitrascu ..................................... 153
Microvascular Doppler ultrasound in children with acute pyelonephritis
G. Choi, B-K. Je, D. Hong, J. Cha ............................................................................................................................................ 161
Comparison of IOTA three-step strategy and logistic regression model LR2 for discriminating between benign
and malignant adnexal masses
J.J, Hidalgo, A. Llueca, I. Zolfaroli, N. Veiga, E. Ortiz, J.L. Alcázar ....................................................................................... 168
The presence of effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum
tendon is a common phenomenon: a single-center, cross sectional study
L. Xue, Y. Zhang, D. Yan, J. Fu, Z. Liu ...................................................................................................................................... 176
Assessment of testes with two-dimensional Shear Wave Elastography in patients with operated inguinal hernia
M.S. Durmaz, F. Ates, S. Arslan, T. Kara, F.G. Durmaz, M.A. Eryilmaz, K. Arslan ................................................................. 181
Congenital complete atrioventricular block from literature to clinical approach – a case series and literature
review
L. Gozar, C. Marginean, A. Fagarasan, I. Muntean, A. Cerghit-Paler, D. Miklosi, R. Toganel ............................................... 188
Reviews
Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors – an updated review of
the literature
S. Stoia, G. Băciuț, M. Lenghel, R. Badea, M. Băciuț, S. Bran, D. Cristian ............................................................................. 194
The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures: a systematic review and
meta-analysis
J. Wu, Y. Wang, Z. Wang ............................................................................................................................................................ 203
Understanding the role of echocardiography in patients with obstructive sleep apnea and right ventricular
subclinical myocardial dysfunction – comparison with other conditions affecting RV deformation
I.M. Chetan, B. Domokos Gergely, A. Albu, R. Tomoaia, D.A. Todea ....................................................................................... 213
Pictorial essay
First trimester fetal heart evaluation. A pictorial essay
I.C. Rotar, D. Mureșan, C. Mărginean, D.G. Iliescu, Ș. Tudorache ......................................................................................... 220
Medical Ultrasonography
Official Journal of the Romanian Society for Ultrasonography in Medicine and Biology
Medical Ultrasonography (formerly Revista Româna de Ultrasonografie from 1999 to 2008) is the official publication of the
Romanian Society for Ultrasonography in Medicine and Biology (SRUMB). Starting with 2008 the entire content of Medical
Ultrasonography is published in English, quarterly. The journal aims to promote ultrasound diagnosis by publishing papers in a
variety of categories, including Original papers, Review Articles, Pictorial Essays, Technical Innovations, Case Report, or Letters to
the Editor (fundamental as well as methodological and educational papers). The published papers cover a wide variety of discipline
of ultrasound. The journal also host information regarding the society’s activities, the scheduling of accredited training courses in
ultrasound diagnosis, as well as the agenda of national and international scientific events.
Medical Ultrasonography is now listed in Science Citation Index Expanded/ ISI Thomson Master Journal List, Medline/
PubMed, Scopus, Pro Quest, Ebsco, and Index Copernicus data bases. Impact Factor 1.553 (JCR 2019); 5 year IF=1.598
Editorial Office
2nd Medical Clinic, 2-4 Clinicilor str., 400006 Cluj-Napoca, Romania
Tel.: +4 0264 591942/442, Fax: +4 0264 596912, Email: medultrasonography@gmail.com
Contact person: Daniela Fodor, email: dfodor@ymail.com
Journal web site: http://www.medultrason.ro
Editorial board
Editor in Chief Methodological adviser Editors Assistant Editors English language editors
Daniela Fodor Petru Adrian Mircea Radu Ion Badea Carolina Solomon Sally Wood-Lamont
Sorin Marian Dudea Bogdan Chis Ioana Robu
Oana Serban
Members
Mihaela Băciuţ (Cluj-Napoca, Romania) Richard Hoppmann (Columbia, South Carolina, USA Alina Popescu (Timişoara, Romania)
Boris Brkljacic (Zagreb, Croatia) Walter Grassi (Ancona, Italy) Alper Ozel (Istambul, Turkey)
Ciprian Brisc (Oradea, Romania) Lucas Greiner (Wuppertal, Germany) Adrian Săftoiu (Craiova, Romania)
Vito Cantisani (Rome, Italy) Norbert Gritzmann (Salzburg, Austria) Paul Singh Sidhu (London, UK)
Anca Ciurea (Cluj-Napoca, Romania) Zoltán Harkányi (Budapest, Hungary) Zeno Spârchez (Cluj-Napoca, Romania)
Sorin Crişan (Cluj-Napoca, Romania) Anamaria Iagnocco (Rome, Italy) Ioan Sporea (Timişoara, Romania)
Adrian Costache (Bucureşti, Romania) Adnan Kabaalioglu (Antalya, Turkey) Florin Stamatian (Cluj-Napoca)
Jarosław Czubak (Otwock, Poland) Daniel Lichtenstein (Paris, France) Dan Stănescu (Bucureşti, Romania)
Christoph Dietrich (Frankfurt am Main, Germany) Carmen Mihaela Mihu (Cluj-Napoca, Romania) Iwona Sudoł-Szopińska (Warsaw, Poland)
Dan Dumitraşcu (Cluj-Napoca) Dan Mihu (Cluj-Napoca, Romania) Kazmierz Szopinski (Warsaw, Poland)
Viorela Enăchescu (Craiova, Romania) Daniel Muresan (Cluj-Napoca, Romania) Adrian Şanta (Sibiu, Romania)
Otilia Fufezan (Cluj-Napoca, Romania) Luca Neri (Milan, Italy) Roxana Sirli (Timişoara, Romania)
Odd Helge Gilja (Bergen, Norway) Monica Platon Lupsor (Cluj-Napoca, Romania)
Tehnical staff Instruction for authors: Subscription information:
Iulia China Full instructions are available online at http://www.medultrason.ro/ Medical Ultrasonography is published quarterly.
authors-guidelines/ and at the end pages of the journal. ISSN (print) 1844–4172; ISSN (online) 2066–8643
The annual subscription:
For Romania – 150.00 lei for individuals For other countries (including postage fees) – 60 euro for individuals
– 350.00 for institutions – 120 euro for institutions
Bank account:
– in lei: SRUMB – Filiala Cluj-Napoca (CIF 10232679), BRD, sucursala Cluj-Napoca, Cont: RO62 BRDE 130S V959 5116 1300
– in euro: SRUMB – Filiala Cluj-Napoca (CIF 10232679), BRD, sucursala Cluj-Napoca, Cont: RO54 BRDE 130S V959 5132 1300
How to subscribe: Subscribers may pay in the bank account already mentioned. Orders should be sent online at www.medultrason.ro
“Subscription” or by email to the Editorial Office (dfodor@ymail.com)
Publishing House: “Iuliu Haţieganu” Medical Publishing House, Cluj-Napoca, Contact person: Ioana Robu,
8, Victor Babeş str, Cluj-Napoca, Tel.: + 40-264-592629 / 590832, Fax: + 40-264-598820, E-mail: irobu@umfcluj.ro
All right reserved to the Medical Publishing House of the “Iuliu Hatieganu” University of Medicine and Pharmacy, Cluj-Napoca;
no parts of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise without the permission of the publisher.
Contents
(continued)
Continuing education
The utility of ultrasound in the diagnostic evaluation of the posterior ankle joint
W. Konarski, T. Poboży .............................................................................................................................................................. 226
Case report
The role of multimodal imaging in the diagnosis of an asymptomatic patient with congenital anomaly
R. Tomoaia, A. Molnar, R.Ș. Beyer, A. Dădârlat-Pop, F. Frîngu, D. Gurzau, G. Simu, I.A. Minciună, B. Caloian,
D. Zdrenghea, D. Pop ................................................................................................................................................................ 231
A rare cause of biliary obstruction – intraductal neuroendocrine tumor of the right hepatic biliary duct:
a case report
M. Danila, R. Sirli, A. Popescu, N. Iacob, A-M. Ghiuchici ....................................................................................................... 235
Department of Gastroenterology and Hepatology, “Victor Babeş” University of Medicine and Pharmacy, Regional
Center of Research in Advanced Hepatology, Academy of Medical Science, WFUMB Center of Education, Timişoara,
Romania
Nonalcoholic Liver Disease (NAFLD) has become authors. Two main categories of patients are considered at
the most frequent hepatological entity in the clinical risk and should be screened for fatty liver and fibrosis: pa-
practice in the developed world. The increasing number tients with type 2 DM and metabolic subjects. Concerning
of overweight and obese subjects, together with the high diabetic subjects, very often they also have a metabolic
prevalence of type 2 diabetes mellitus (DM) and meta- syndrome and are obese, increasing the risk of liver dis-
bolic syndrome, results in this condition being present in ease. In a Dutch study [4] performed in a cohort of 3,041
approximately 25% of the population. A new terminol- participants (mean age 66.0 years), 5.6% of participants
ogy that tends to replace the term NAFLD is MAFLD had significant fibrosis (liver stiffness measurements by
(Metabolic Associated Fatty Liver Disease) which seems Transient Elastography ≥8.0 kPa). In multivariate analysis
to show better the connection between fatty liver and the age, alanine aminotransferase, spleen size and the com-
spectrum of metabolic syndrome [1,2]. bined presence of diabetes mellitus and ultrasound stea-
For a long time, fatty liver in obese or type 2 DM tosis (OR, 5.20) were associated with significant fibrosis.
patients was considered a benign disease and not much In a study performed by our group [5] in a cohort
attention was paid to this entity. In the last years, many of 776 type 2 DM patients, evaluated with TE and CAP
papers have shown that it can progress to significant fi- (Controlled Attenuation Parameter), we found that 60.3%
brosis or cirrhosis in a proportion of subjects. of them had severe steatosis and 19.4% had advanced fi-
Now the question arises to screen or not to screen the brosis. In a review paper [6] results from two epidemiol-
general population for fatty liver and the methodology in ogy reports from United States documented steatosis in
how to carry it out? A populational study performed in 70-74% of unselected patients with type 2 DM, advanced
Korea using Magnetic Resonance Elastography (MRE) liver fibrosis in 6%-15%, and previously unrecognized
in a large cohort of 8183 subjects, found that 9.5% of cirrhosis in 3-8%.
the patients had at least significant fibrosis (F≥2), which All these facts are arguments for screening for stea-
included 2.6% subjects with at least advanced fibrosis tosis and fibrosis in such categories of patients. But how
(F≥3) [3]. In this study, multivariate analysis revealed should this screening be performed? Simple and inexpen-
that liver fibrosis was associated with age, male sex, type sive biological tests such as FIB4 are useful for the gen-
2 DM, abnormal liver function test, HBsAg positivity eral practitioner for fibrosis assessment. However, what
and obesity, as well as with metabolic syndrome. How- we propose is to use ultrasound and ultrasound based
ever, at this moment, screening of the general population elastography for steatosis and fibrosis assessment. The
is difficult to perform since it would comprise a huge co- accuracy of standard ultrasound for a semi-quantitative
hort of individuals. assessment of liver steatosis is quite good (especially for
On the other hand, screening the population at risk moderate or severe steatosis), with an AUROC of 0.93
has been proposed more and more frequently by different [7]. Using Transient Elastography with CAP for the as-
sessment of fibrosis and steatosis, the AUROC’s range
Received Accepted from 0.80 to 0.90 for the detection of significant steatosis
Med Ultrason and fibrosis, respectively.
2021, Vol. 23, No 2, 133-134 In the last few years advanced ultrasound systems
Corresponding author: Prof. Ioan Sporea, MD, PhD
13, Snagov Street,
have been developed that can perform ultrasound based
300482 Timisoara, Romania elastography and also steatosis quantification. Some of
E-mail: isporea@umft.ro these new systems can also evaluate the viscoelastic
134 Ioan Sporea To screen or not to screen for NAFLD?
properties of the liver tissue, which are considered to be ture for Metabolic Associated Fatty Liver Disease. Gastro-
a measure of inflammation. Thus, multiparametric ultra- enterology 2020;158:1999-2014.e1.
sound systems are now available in the clinical practice 2. Lin S, Huang J, Wang M, et al. Comparison of MAFLD
and NAFLD diagnostic criteria in real world. Liver Int
for the assessment of diffuse liver disease. Nowadays,
2020;40:2082-2089.
modules for steatosis and fibrosis assessment are imple-
3. Nah EH, Cho S, Kim S, Chu J, Kwon E, Cho HI. Preva-
mented by some companies, in middle range ultrasound lence of liver fibrosis and associated risk factors in the
machines (not only in high end systems). Korean general population: a retrospective cross-sectional
With the implementation of “point of care” ultra- study. BMJ Open 2021;11:e046529.
sound, performed by specialists in their consultation 4. Koehler EM, Plompen EP, Schouten JN, et al. Presence of
room, the evaluation of MAFLD patients has become diabetes mellitus and steatosis is associated with liver stiff-
more feasible and in a very short time. Regarding the ness in a general population: The Rotterdam study. Hepa-
cost-utility and cost-effectiveness of using ultrasound tology 2016;63:138–147.
and ultrasound based elastography for the screening of 5. Sporea I, Mare R, Popescu A, et al. Screening for Liver
NAFLD patients, some studies have shown positive re- Fibrosis and Steatosis in a Large Cohort of Patients with
Type 2 Diabetes Using Vibration Controlled Transient Elas-
sults [8,9].
tography and Controlled Attenuation Parameter in a Single-
Thus, finally, to answer the question from the title, Center Real-Life Experience. J Clin Med 2020;9:1032-
I believe that this is the perfect time to start screening 1045.
patients for NAFLD, in order to evidence this disease in 6. Zoler ML. Widespread Liver Disease Missed in Patients
the early stages [10]. Using ultrasound machines and ul- With Type 2 Diabetes. Medsearch 2021 Jan 25. Available at:
trasound based elastographic methods, we can undertake https://www.medsearchuk.com/widespread-liver-disease-
this screening with good results and with a low cost. Not missed-in-patients-with-type-2-diabetes-widespread-liver-
only screening in NAFLD patients should be proposed, disease-missed-in-patients-with-type-2-diabetes/
but also in other categories of patients, such as alcoholic 7. Hernaez R, Lazo M, Bonekamp S, et al. Diagnostic Accura-
subjects (for alcoholic liver disease: ALD). And if this cy and Reliability of Ultrasonography for the Detection of
Fatty Liver: A Meta-Analysis. Hepatology 2011;54:1082–
screening can be performed as a “point of care” method,
1090.
it is a better solution, both for patients and physicians.
8. Zhang E, Wartelle-Bladou C, Lepanto L, Lachaine J, Clout-
Thus, my advice is to start screening such patients ier G, Tang A. Cost-utility analysis of nonalcoholic steato-
with steatosis and fibrosis assessment using modern ul- hepatitis screening. Eur Radiol 2015;25:3282-3294.
trasound systems, for an early diagnostic and a therapeu- 9. Tapper EB, Sengupta N, Nunink MG, Afdhal NH, Lai
tic solution. M. Cost-Effective Evaluation of Nonalcoholic Fatty Liv-
er Disease With NAFLD Fibrosis Score and Vibration
References Controlled Transient Elastography. Am J Gastroenterol
2015;110:1298-1304.
1. Eslam M, Sanyal AJ, George J; International Consensus 10. Sporea I, Șirli R. Nonalcoholic Fatty Liver Disease and the
Panel. MAFLD: A Consensus-Driven Proposed Nomencla- Need for Action. J Gastrointestin Liver Dis 2020;29:139-141.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 135-139
DOI: 10.11152/mu-2746
1Research Center of Gastroenterology and Hepatology, University of Medicine and Pharmacy Craiova, 2Faculty
of Automation, Computers and Electronics, University of Craiova, Craiova, 3Faculty of Mechanics, University of
Craiova, Craiova, Romania
Abstract
Aim: In this paper we proposed different architectures of convolutional neural network (CNN) to classify fatty liver
disease in images using only pixels and diagnosis labels as input. We trained and validated our models using a dataset of 629
images consisting of 2 types of liver images, normal and liver steatosis. Material and methods: We assessed two pre-trained
models of convolutional neural networks, Inception-v3 and VGG-16 using fine-tuning. Both models were pre-trained on
ImageNet dataset to extract features from B-mode ultrasound liver images. The results obtained through these methods were
compared for selecting the predictive model with the best performance metrics. We trained the two models using a dataset of
262 images of liver steatosis and 234 images of normal liver. We assessed the models using a dataset of 70 liver steatosis im-
ages and 63 normal liver images. Results. The proposed model that used Inception v3 obtained a 93.23% test accuracy with a
sensitivity of 89.9%% and a precision of 96.6%, and areas under each receiver operating characteristic curves (ROC AUC) of
0.93. The other proposed model that used VGG-16, obtained a 90.77% test accuracy with a sensitivity of 88.9% and a precision
of 92.85%, and areas under each receiver operating characteristic curves (ROC AUC) of 0.91. Conclusion. The deep learning
algorithms that we proposed to detect steatosis and classify the images in normal and fatty liver images, yields an excellent
test performance of over 90%. However, future larger studies are required in order to establish how these algorithms can be
implemented in a clinical setting.
Keywords: fatty liver disease; NAFLD; convolutional neural networks; deep learning
Introduction ease (FLD) has become the most common liver disease
in Western countries, non-alcoholic fatty liver disease
With the rapid growth of the population and the in- (NAFLD) affecting around 25% of the global adult pop-
creasing rates of obesity and diabetes, fatty liver dis- ulation [1]. Liver steatosis is defined as the accumula-
tion of more than 5% of fat in the hepatocytes [2]. In
Received 22.08.2020 Accepted 01.12.2020 the natural evolution of NAFLD, if no early strategies
Med Ultrason are adopted, the disease can progress to steatohepatitis
2021, Vol. 23, No 2, 135-139 (NASH) which is a risk factor for liver fibrosis, cirrhosis
Corresponding author: Adrian Săftoiu MD PhD MSc
Professor of Diagnostic and Therapeutic
and even hepatocellular carcinoma [3]. In order to con-
Techniques in Gastroenterology firm the diagnosis, to quantify liver steatosis and stage
University of Medicine and Pharmacy NAFLD, liver biopsy is still considered the gold stand-
Craiova, Romania ard. However, it is an invasive procedure which carries a
Phone: +40 744 823355
Fax: +40 251 310287
high risk of serious complications and, therefore, it can-
E-mail: a drian.saftoiu@umfcv.ro not be routinely used for the evaluation and follow-up of
adriansaftoiu@gmail.com NAFLD [4].
136 Elena Codruța Constantinescu et al Transfer learning with pre-trained deep convolutional neural networks
Non-invasive imaging methods used for the diagno- Materials and methods
sis of FLD include ultrasonography, computed tomogra-
phy, magnetic resonance and magnetic resonance-based Pre-trained convolutional neural network
fat quantification techniques. Conventional B-mode ul- implementation and parameter details
trasonography is mostly preferred for screening and for Very deep networks are prone to overfit and large
the initial assessment of FLD due to the wide availabil- convolution operations are computationally expensive.
ity and low costs [5]. Although a normal ultrasound im- The Inception network, on the other hand, is complex
age cannot rule out liver steatosis, the overall sensitivity and uses many improvements to obtain better perfor-
and specificity of ultrasound are comparable to those of mance, both in terms of speed and accuracy. VGG16,
histology, increasing with the degree of steatosis [5-7]. another CNN architecture which is considered to be an
However, FLD can be diagnosed on ultrasound based excellent convolution neural model, is a large network,
on some suggestive parameters: diffuse hyperechoic comprising approximately 138 million parameters [18].
structure (“bright liver”), deep beam attenuation, liver- In this study, we used fine-tuning which is a type
to-kidney contrast, bright vessel walls [8,9]. Taking into of transfer learning. We applied fine-tuning to the pre-
consideration the public health burden given by NAFLD, trained DL models Inception v3 and VGG-16. The top
there is a great need for the development of a non-inva- layers of the pre-trained models are replaced by new fully
sive approach for the correct diagnosis and to determine connected layers with random parameters, while the lay-
the progression of the liver disease in order to establish ers below are kept frozen. The new fully connected layers
an early treatment. can learn patterns from previous learned convolutional
Over time, in order to improve the diagnostic accu- layers, because a very small learning rate is utilised. By
racy of FLD and to make a quantitative assessment of applying fine-tuning, liver steatosis could be recognised
steatosis, several computer-aided methods were inves- even if our networks, Inception v3 and VGG-16 were not
tigated, including hepato-renal index (HRI), gray level originally trained. This method achieved higher accuracy
co-occurrence matrix (GLCM), artificial intelligence than feature extraction [17,18]. After rescaling the im-
(AI) methods such as traditional machine learning (ML) ages and using image augmentation, we flowed them in
algorithms and deep learning (DL) algorithms. One of batches of 16. In order to reduce overfitting and improve
the most popular types of DL algorithms and widely used the model’s ability to generalize, we used dropout, activ-
in medical imaging are convolutional neural networks ity regularization and kernel regularization (see: https://
(CNN). CNN consist of a series of layers (convolutional, github.com/keras-team/keras).
pooling and fully connected) that can perform end-to-end The generic architecture of the proposed methods
supervised learning in order to identify patterns, lines or based on transfer learning with fine-tuning can be ob-
edges while some of the layers that are ”hidden” can served in figure 1.
conduct unsupervised learning tasks. However, CNN ar- Inception v3 model
chitecture consists of many convolutional layers, which We used Inception v3 model pre-trained on ImageNet
means that a great deal of data is needed for the train- dataset to extract high-level features from B-mode ultra-
ing set in order to acquire competent accuracy [10]. To sound liver images [17]. Our proposed model included
overcome this issue, data augmentation techniques have all the layers in the Inception v3 network, except for the
been proposed to reduce overfitting. In order to increase last fully connected layer, making all the layers non-
the efficiency of CNN, transfer learning has become very trainable in order to reduce the overfitting. At the top of
popular because it considerably reduces the training time
and requires less data to train, while increasing the classi-
fication performance. Up to date, the research in clinical
application of AI in liver imaging is small and limited,
and more studies are required to demonstrate the effi-
ciency of DL algorithms [11] but the importance of CNN
in the detection of liver masses has been demonstrated in
previous studies [12-14]. Moreover, several small studies
used DL with B-mode ultrasound images for the diagno-
sis and classification of FLD [15,16].
The aim of our study was to analyse the test perfor-
mance of recent DL algorithms for the ultrasound evalu- Fig 1. The generic representation of the transfer learning meth-
ation of liver steatosis. od used to detect liver steatosis
Med Ultrason 2021; 23(2): 135-139 137
the network we added two fully connected layers, one
with 1024 hidden units and ReLU activation and the oth-
er was the last sigmoid dense layer for classification. Our
optimizer was RMSprop with a learning rate of 0.0001
(see: https://github.com/keras-team/keras).
VGG-16 model
Another model that we used was the VGG-16 net-
work pre-trained on ImageNet [18]. The model we used
included all the layers in the network, except for the last
fully connected layer. At the top of the network we added
three fully connected layers with 512, 128, and 64 hidden
units and ReLU activation. The last layer is the sigmoid
dense layer used for classification. Our optimizer was
Adam with a learning rate of 0.0001 [19].
Patient selection and image dataset Fig 2. The selection of non-overlapping patches from images
The dataset contains images from 60 patients, 30 pa-
tients with hepatic steatosis and 30 healthy subjects, both and 234 normal liver images. For the test dataset we used
male and female, age ranging from 18 to 92. The partici- 70 liver steatosis images and 63 normal liver images.
pants were selected by an expert hepatologist from the Statistical analysis
University of Medicine and Pharmacy, Research Center The medical diagnosis metrics used to analyse the
of Gastroenterology and Hepatology, Craiova from the performance of our CNN models were: sensitivity, preci-
outpatient clinic of a private healthcare network. Exclu- sion, test accuracy and area under curve (AUC).
sion criteria included excessive alcohol consumption or The other metrics used to evaluate the classifier qual-
>20 g ethanol per day, history or clinical and/or laborato- ity were the receiver operating characteristic (ROC) met-
ry evidence of liver disease, hepatotoxic medication use. ric and Precision-Recall (PR). While precision measures
Written informed consent was obtained from all patients. the relevancy of the result, recall measures how many
The study is non-interventional and all procedures are relevant results are returned.
used in the current daily practice.
The diagnosis of NAFLD was based on the imag- Results
ing studies (B-mode ultrasound and ARFI elastography)
along with the clinical exclusion of the patients who con- We used two pre-trained CNN models to detect the
sume > 20 g ethanol per day. The B-mode ultrasound characteristic features in each liver image. Our results
images were collected by the same physician and all show a final diagnosis accuracy of 93.23% and the AUC
examinations were performed on Siemens Acuson NX3 of 0.93 for the Inception v3 model. By comparison, when
ultrasound machine. The ultrasound diagnosis of steato- using the VGG-16 model, the final diagnosis accuracy
sis was indicated by diffuse hyperechoic structure, deep was 90.77% and the area under curve of 0.91. All metrics
beam attenuation and increased liver echogenicity in computed for Inception v3 vs. VGG-16 are summarized
relation to the right kidney. Participants were evaluated in Table I and showed the improvements achieved by us-
with pSWE, an alternative method for the non-invasive ing the Inception v3 model.
evaluation of liver fibrosis that uses acoustic radiation The ROC curves for the two models (Inception v3 vs.
force impulse imaging (ARFI) to induce tissue disloca- VGG-16) show the rate of false positive is near to zero
tion. The examinations were performed by another phy- while the rate of true positive is between 0.9 and 1 (fig
sician on a Siemens Acuson S2000 machine. All the par- 3a). The precision-recall curves of the two models (In-
ticipants had ten valid measurements performed in the ception v3 vs. VGG-16) demonstrate a better precision
right liver lobe with IQR/M ≤30% and a median value and recall of the Inception v3 model (fig 3b). High scores
ranging from to 1.8 to 3.7 kPa.
The database was created by cropping non-overlap- Table I. Evaluation metrics for the classification of non-alco-
ping patches (fig 2) and by specific transformation such holic fatty liver disease using the two pre-trained CNN models.
as resizing and shifting, obtaining 629 grayscale liver im- Accuracy AUC Sensitivity Precision
ages with a resolution of 75x75. Out of the 629 images,
Inception v3 93.23% 0.93 88.9% 96.6%
496 were used for training and 133 for testing. For the
training dataset, we used 262 images with liver steatosis VGG-16 90.77% 0.91 88.9% 91.8%
138 Elena Codruța Constantinescu et al Transfer learning with pre-trained deep convolutional neural networks
1Department of Gastroenterology and Hepatology, “Victor Babeş” University of Medicine and Pharmacy Timişoara,
2Research Center of Gastroenterology and Hepatology, University of Medicine and Pharmacy Craiova, 3Regional
Institute of Gastroenterology and Hepatology “Prof. Dr. Octavian Fodor”, ”Iuliu Hațieganu” University of Medicine
and Pharmacy Cluj Napoca, 4Department of Gastroenterology, “Gr. T. Popa” University of Medicine and Pharmacy
Iaşi, 5Center of Internal Medicine, Fundeni Clinical Institute, “Carol Davila” University of Medicine and Pharmacy,
Bucharest, 6Ponderas and Neolife Hospitals, Bucharest, 7Department of Anatomy and Embryology, “Victor Babeş”
University of Medicine and Pharmacy, Timişoara, 8Department of Gastroenterology, University of Oradea, Romania
Abstract
Aim: Contrast-enhanced ultrasound (CEUS) has become a relevant imaging method for the evaluation of focal liver le-
sions (FLL). The aim of this study was to evaluate the performance of CEUS for the assessment of focal nodular hyperplasia
(FNH) in a large study group. Material and methods: We performed a multicentre prospective observational study, which
included successive CEUS examinations from fourteen Romanian centres. CEUS examinations were performed in de novo
FLL, using low mechanical index ultrasound, following an intravenous bolus of 2.4 ml SonoVue. CEUS was considered
conclusive for FNH if a typical pattern was present following contrast (rapid “spoke-wheel” enhancement during the arterial
phase, hyperenhanced lesion during venous phase, hyper- or isoenhanced in the late phase). In all cases a reference method was
available (contrast enhanced CT or MRI or biopsy). The trial was registered in clinicaltrials.gov (Identifier NCT01329458).
Results: During the 6 years study, 2062 “de novo” FLL were evaluated by CEUS. From this cohort, 94/2062 (4.5%) had a
typical enhancing pattern for FNH as described in the EFSUMB guidelines. Contrast enhanced CT/MRI and biopsy diagnosed
additional 15 FNH. From the 94 cases diagnosed as FNH by CEUS, in nine the final diagnosis was different (five of them
adenomas). CEUS had 85% sensitivity, 99.5% specificity, 90.4% positive predictive value, 99.2% negative predictive value
and 98.8% diagnostic accuracy for the diagnosis of FNH. Conclusions: CEUS is a sensitive and very specific method for the
diagnosis of FNH.
Keywords: Contrast Enhanced Ultrasound; focal liver lesions; focal nodular hyperplasia; multicentre study
Fig 1. a) Slightly hypoechoic FLL in the right liver lobe. b) The FLL shows “spoke-wheel” pattern on Doppler US. c) CEUS, arterial
phase – the lesion is completely hyperenhanced 11 seconds after bolus injection – excentric feeding artery also visible. d) CEUS,
portal phase – the lesion is slightly hyperenhancing with visible central scar. e) CEUS, late phase – the lesion is iso/hyperenhancing
with well visible central scar (arrow).
per standard protocol [6,12]. Lesions’ enhancement pat- nous phase, hyper or isoenhanced lesion in the late phase,
terns were studied in 3 phases: arterial (10-30 seconds sometimes with the visualization of a central hypoechoic
following contrast bolus), portal (30-120 seconds) and scar (fig 1). A CEUS examination was considered con-
late phase (>120 seconds) [6,12]. In each patient, the clusive if the FLL had a typical enhancement pattern ac-
contrast study duration was at least 5 minutes after bo- cording to national and international guidelines [6,12]
lus injection. To document the study, four video files no and inconclusive if not. The CEUS diagnosis was com-
longer than 30 seconds each were captured, containing pared with the final diagnosis established based on all
conventional B-mode examination, the arterial phase, the available imaging and clinical data: contrast enhanced
portal phase and the late phase. CT, and/or MRI, and/or histology.
CEUS vascular patterns were defined by compar- Statistical analysis
ing the FLL’s enhancement pattern to the surrounding Statistical analysis was performed using the Graph-
liver parenchyma. We classified CEUS vascular pat- Pad Prism program, version 7.02 (GraphPad Software,
terns as: homogeneous hyperenhancement (the whole La Jolla, USA). We assessed the accuracy of CEUS for
FLL showed homogeneous enhancement); heterogene- FNH characterization. The sensitivity (Se) was calcu-
ous hyperenhancement (mixed irregular areas of contrast lated as true positive cases divided by the total number
enhancement were present); rim-like hyperenhancement of cases in which the disease was present; the specificity
(a peripheral hyperenhancement limited to less than 25% (Sp) was calculated as true negative cases divided by the
of the tumour); iso-enhancement (the FLL enhanced the total number of cases in which the disease was absent;
same as the adjacent parenchyma from the same depth); the positive predictive value (PPV) was calculated as true
hypoenhancement (the lesion enhanced less than the positive cases divided by all CEUS positive cases; the
adjacent parenchyma at the same depth); wash-out (hy- negative predictive value (NPV) was calculated as true
poenhancement in the portal or late phases preceded by negative cases divided by all CEUS negative cases and
hyper or iso-enhancement in the arterial phase). accuracy was calculated as the sum of true positive and
A CEUS diagnosis of FNH was established after the true negative cases divided by the total number of cases.
contrast study based on the patterns described in guide- We included in the statistical analysis all cases reported;
lines [6,12]: rapid “spoke-wheel” enhancement during we considered the inconclusive CEUS cases as wrongly
the arterial phase, hyperenhanced lesion during the ve- diagnosed.
Med Ultrason 2021; 23(2): 140-146 143
Results Table I. Patients’ characteristics
Number of patients 2062
During the 6 years study, 2062 “de novo” FLL were Mean age (years) 52.4±7.5
evaluated by CEUS. Patients’ characteristics are present- Gender, women 914 (44.3)
ed in Table I.
Underlying chronic liver disease 727 (35.3%)
From this cohort, 94/2062 (4.5%) had a typical en-
CEUS diagnosis
hancing pattern for FNH. Contrast enhanced CT/MRI Hepatocellular carcinomas 685 (33.2)
and biopsy diagnosed additional 15 FNH (12 in cases Hemangiomas 452 (21.9)
labelled as inconclusive on CEUS and one each labelled Metastases 418 (20.3)
by CEUS as haemangioma, adenoma and metastasis). Focal nodular hyperplasias 94 (4.5)
From the 94 cases diagnosed as FNH by CEUS, in nine Regenerative nodules 84 (4.1)
Focal fatty infiltrations 70 (3.4)
cases the final diagnosis was different (five of them ad- Cholangiocarcinomas 57 (2.7)
enomas, two of them hepatocellular carcinomas, one Abscesses 45 (2.2)
haemangioma and one focal fatty infiltration). Thus, the Complex Cysts 43 (2.1)
final diagnosis was FNH in 100 of the 2062 (4.8%) cases. Other benign lesions* 37 (1.8)
The characteristics of patients with FNH are presented in Adenomas 32 (1.6)
Fatty free areas 26 (1.3)
Table II. Other malignant lesions** 19 (0.9)
The indication for CEUS was in most FNH cases an The results are expressed as number, number (%) or mean±SD. *
incidentally found FLL 98% (98/100), in 1% (1/100) pseudolesion, scar area, angiomyolipoma, hamartoma; ** lympho-
case it was the evaluation of a patient with known onco- ma, hemangiosarcoma, hepatic epithelioid hemangioendothelioma.
logic disease and in 1% (1/100) case it was an inconclu-
sive CE-CT scan. At the moment of CEUS examination, Table II. The characteristics of patients with focal nodular
none of the patients diagnosed by CEUS with FNH were hyperplasia (FNH)
known with underlying liver disease. However, two of Number of patients with FNH 100
them were later diagnosed with underlying cirrhosis and Mean age 40.4±13.7 years
2 with chronic hepatitis. Gender, women 75 (75)
CE-MRI was the gold standard in 30% (30/100) cas- Underlying chronic liver disease 4 (4)
es. In 6% (6/100) patients histology was available – one Size:
of them after surgery performed for HCC. ≤ 35 mm 39 (39)
By comparing CEUS with the reference method (ei- ≥ 36 mm 61 (61)
ther CE-CT, CE-MRI or histology) CEUS had 85% Se, The results are expressed as number (%) or mean±SD
99.5% Sp, 90.4% PPV, 99.2% NPV and 98.8% diagnos-
tic accuracy for the diagnosis of FNH. hand, CEUS can be performed immediately after a con-
ventional B mode US examination and thus a diagnosis
Discussion can be established in minutes.
CEUS is a real-time imaging technique, which, simi-
Focal liver lesions (FLLs) are frequently discovered lar to contrast enhanced CT or MRI, shows tissue perfu-
during conventional B mode ultrasound (US) examina- sion, and is able to identify the type of FLL based on the
tion, either during a routine scan for various abdomi- enhancement pattern in arterial, portal and late vascular
nal symptoms or during the follow-up of chronic liver phases. The value of CEUS for the differential diagnosis
disease or of oncologic patients. In most cases, only B of FLLs was proved in a German national multicentre
mode US is not sufficient for a definite diagnosis. This study (DEGUM study) [13] and in a French national
is particularly true when FNH is involved, since its ap- multicentre study (STIC study) [14], each including more
pearance is totally uncharacteristic, sometimes isoechoic, than 1000 FLL and later in a Romanian multicentre study
or slightly hyper- or hypoechoic, the most suggestive for [15]. Meta-analyses demonstrated that CEUS accuracy
the diagnosis being the central scar, which is visible only is similar to contrast enhanced CT and MRI [16-18].
in 70-80% of cases [8,10,11]. Thus, a contrast imaging Other advantages of CEUS as compared with CT and
method or a histopathological exam are required to reach MRI are the absence of radiation exposure (as in CT ex-
a final diagnosis. The time needed for scheduling, and amination) and of side effects (no nephrotoxicity, no io-
the costs of contrast enhanced CT or contrast enhanced dine exposure), and the lower costs [19,20]. Limitations
MRI are not negligible, the same is true regarding guided of CEUS are linked to the limitations of US (poor acous-
biopsy, which is also an invasive procedure. On the other tic window, uncooperative patients) and to the lesion’s
144 Roxana Şirli et al CEUS for the assessment of focal nodular hyperplasia – results of a multicentre study
depth, since FLL located at more than 9 cm from the skin curacy (98.8%) and a slightly lower Se (85%). Regard-
are difficult to examine due to attenuation [6,12]. How- ing the misdiagnosed cases, most confusions were made
ever, if we compare the methods, we discover from the with adenomas. Five cases diagnosed by CEUS as FNH
study of Bröker et al [21] that in a direct comparison be- were in fact adenomas and one case diagnosed by CEUS
tween MRI with liver specific contrast and CEUS, MRI as adenoma was in fact FNH. The confusion is probably
performed better than CEUS for the diagnosis of FNH understandable since both types of FLL show rapid hy-
and hepatocellular adenoma, CEUS having a sensitivity perenhancement in the arterial phase, but in adenoma the
and specificity of 85% and 87%, positive predictive value fill-in occurs initially at the periphery with very rapid
(PPV) of 79% and negative predictive value (NPV) of centripetal filling, as opposed to the centrifuge filling
90%, while MRI reached 95% sensitivity, 95% specific- seen in FNH.
ity, 92% PPV and 97% NPV. Arterial phase can be better evaluated using paramet-
The first guidelines regarding the use of CEUS were ric arrival time (PAT), a post processing modality of eval-
issued in 2004 by the European Federation of Societies uation. In the portal phase, adenoma usually becomes
for Ultrasound in Medicine and Biology (EFSUMB) isoechoic or, more rarely, remains slightly hyperechoic.
[22], which were revised in 2008 [23], in 2012 [6] and in Sometimes, in adenomas the washout occurs in the late
2020 [24], the last two developed in cooperation with the phase, thus being false positive for malignancy [6,12].
World Federation for Ultrasound in Medicine and Biol- The imaging diagnosis of adenoma is very difficult and
ogy (WFUMB), thus with universal validity. Regarding often guided biopsy is needed. In the DEGUM study,
the dose of SonoVue, guidelines recommend the use of CEUS correctly diagnosed 57.9% of the adenomas [13].
2.4 ml (1/2 vial) for the characterization of FLL, even if Computer-assisted diagnosis can also be used to improve
in the last years, with the development of more perfor- the accuracy of differential diagnosis by CEUS between
mant US machines, doses as low as 1.2 ml are frequently FNH and adenoma [30].
used [25]. A too high dose can lead to apparition of arti- In the Roche et al study [31] were included 43 FNH
facts (shadowing, over-enhancement of small structures and 20 adenomas, most lesions diagnosed based on the
and signal saturation, as well as “drowning” of small le- histopathological exam. The conclusion of this study
sions in the signal of adjacent structures). A too low dose was that CEUS had excellent Sp for diagnosing FNH
can lead to an insufficient quantity of microbubbles in (100%), but the Se varies according to the lesion’s size:
the late phases, so that the wash-out cannot be observed. 93% for lesions ≤ 35 mm and 7.7% for lesions > 35 mm,
However, if the dose is not optimal, the injection can be the overall Se being 67.4 %. The authors explained this
repeated, with a higher or lower dose, as needed [25]. observation by the fact that larger FNH have an increased
Furthermore, to better evaluate the wash-out, insonation vascular supply and several feeding arteries and thus,
can be avoided in the arterial and portal phase of the rein- the typical spoke-wheel pattern is not visible. The same
jection, especially useful for diagnosis of hepatocellular observation, that FNHs with signs of centrifugal filling
carcinoma [26]. were smaller than those without the sign (3.1±1.5 cm vs.
According to published data, CEUS is an accurate 5.2±3.2 cm, p=0.000) was made in the study by Wang
method to characterise FNH. In the DEGUM study the et al [29]. In our study, the five FNH misdiagnosed as
accuracy for FNH was 95.5% [27], with 57.1% Se and adenomas were 2.5 cm, 6 cm, 6 cm, 3 cm, 7.5 and 3.2 cm
99.3% Sp. In the STIC study, CEUS had 82.5% Se and in diameter, respectively.
94.3% Sp for the diagnosis of FNH [14]. The calculated Even if both adenoma and FNH are benign lesions, it
pooled sensitivity for FNH was 88% in the Friedrich- is important to be differentiated since in adenoma there is
Rust et al meta-analysis [18] and the Sp can go as high a risk of rupture (and subsequent bleeding) or malignant
as 100% [14,28]. In another study that evaluated the ac- transformation.
curacy of CEUS performed by two operators for the di- MRI with liver specific-contrast seems to outper-
agnosis of 85 biopsy proven FNH, CEUS had 80.9% Se, form CEUS in differentiating FNH form hepatocellular
95.7% Sp, 95.0% PPV, 83.3% NPV and 88.3% accuracy adenoma. In a study by Grazioli et al, MRI had 96.9%
for operator one and 78.7% Se, 93.6% Sp, 92.5% PPV, Se, 100% Sp, 100% PPV, 96.4% NPV, and 98.3% overall
81.5% NPV and 86.2% accuracy for operator two. This accuracy [32].
study concluded that CEUS performed better than con- Two FLL diagnosed by CEUS as FNH proved to be
trast enhanced CT for characterizing dynamic centrifugal hepatocellular carcinomas (HCCs). Both types of lesions
filling or the “spoke-wheel” sign in small lesions [29]. have arterial hyperenhancement but in HCC there is a
In our study, the performance of CEUS to diagnose chaotic pattern [6,12]. Usually, HCCs show mild, late, or
FNH was very good with excellent Sp (99.5%) and ac- very late washout, correlated with the differentiation of
Med Ultrason 2021; 23(2): 140-146 145
the tumour [33,34]. Well-differentiated HCCs show very 3. Gaiani S, Celli N, Piscaglia F, et al. Usefulness of contrast-
late or no washout, thus the CEUS examination should enhanced perfusional sonography in the assessment of
take at least 4-5 minutes [35]. Thus, in well-differenti- hepatocellular carcinoma hypervascular at spiral computed
tomography. J Hepatol 2004;41:421-426.
ated HCCs, the CEUS aspect can be similar to that of
4. Wanless IR, Albrecht S, Bilbao J, et al. Multiple focal nodu-
FNH but there is a very important factor that should be
lar hyperplasia of the liver associated with vascular malfor-
taken into consideration: up to 90% of HCCs occur on mations of various organs and neoplasia of the brain: a new
a background of chronic liver disease with severe fibro- syndrome. Mod Pathol 1989;2:456-462.
sis and cirrhosis [36]. US based elastography, available 5. Dietrich CF, Schuessler G, Trojan J, Fellbaum C, Ignee A.
in most US machines able to perform CEUS, should be Differentiation of focal nodular hyperplasia and hepatocel-
used to rule in or rule out the presence of cirrhosis in a lular adenoma by contrast-enhanced ultrasound. Br J Ra-
point of care US evaluation [37]. Thus, it is important diol 2005;78:704-707.
to keep in mind that a new lesion on a cirrhotic liver, 6. Claudon M, Dietrich CF, Choi BI, et al. Guidelines and
hyperenhancing in the arterial phase on CEUS, is prob- good clinical practice recommendations for Contrast En-
ably HCC [35,38]. Computer assisted diagnosis [39] as hanced Ultrasound (CEUS) in the liver - update 2012: A
WFUMB-EFSUMB initiative in cooperation with repre-
well as parametric imaging could be used to improve the
sentatives of AFSUMB, AIUM, ASUM, FLAUS and ICUS.
differential diagnosis by CEUS between FNH and HCC, Ultrasound Med Biol 2013;39:187-210.
the mean transit time being significantly longer in FNH 7. Piscaglia F, Lencioni R, Sagrini E, et al. Characterization of
than in HCC [40]. focal liver lesions with contrast-enhanced ultrasound. Ul-
Both HCCs misdiagnosed as FNH were in patients trasound Med Biol 2010;36:531-550.
with incidentally discovered FLL, who were not known 8. Carlson SK, Johnson CD, Bender CE, Welch TJ. CT of fo-
with chronic liver disease, but in whom, starting from the cal nodular hyperplasia of the liver. AJR Am J Roentgenol
HCC, compensated liver cirrhosis was diagnosed. 2000;174:705-712.
A limitation of or study is that the pathological exam 9. Ricci P, Laghi A, Cantisani V, et al. Contrast-enhanced so-
was available only in a small number of cases (only 7 nography with SonoVue: enhancement patterns of benign
focal liver lesions and correlation with dynamic gado-
of the 100 FNH), but an advantage is the large number
benate dimeglumine-enhanced MRI. AJR Am J Roentgenol
included and the prospective design of the study.
2005;184:821-827.
In conclusion, CEUS is an accurate method to diag- 10. Irie H, Honda H, Kaneko K, et al. MR imaging of focal
nose FNH, the main difficulties occurring in differentiat- nodular hyperplasia of the liver: value of contrast-enhanced
ing FNH from adenomas, especially in large lesions. dynamic study. Radiat Med 1997;15:29-35.
11. D’Onofrio M, Crosara S, De Robertis R, Canestrini S, Mu-
Acknowledgements celli RP. Contrast-Enhanced Ultrasound of Focal Liver Le-
sions. AJR Am J Roentgenol 2015;205:W56-W66.
Part of the research published in this paper was made 12. Sporea I, Badea R, Brisc C, et al. Romanian National
with support from the grant awarded by the “Victor Guidelines on Contrast Enhanced Ultrasound in clinical
Babeş” University of Medicine and Pharmacy Timişoara, practice. Med Ultrason 2017;19:401-415.
in PROGRAMUL III – C2 – PCFI - 2015/2016. 13. Strobel D, Seitz K, Blank W, et al. Contrast-enhanced ul-
The authors wish to acknowledge the help provided trasound for the characterization of focal liver lesions--di-
by the following colleagues: Dorina Pestroiu Calescu, agnostic accuracy in clinical practice (DEGUM multicenter
trial). Ultraschall Med 2008;29:499-505.
Lucian Ciobâca, Liana Gheorghe, Mihai Socaciu, Alina
14. Tranquart F LGA, Correas JM, Ladam Marcus V, et al. Role
Martie, Siegfried Christian Ivasc, Attila Tamas, Costin of contrast-enhanced ultrasound in the blinded assessment
Theodor Streba, Mihaela Iordache, Iulia Simionov, Mari- of focal lesions in comparison with MDCT and CEMRI:
ana Jinga, Adrian Anghel, Silviu Marcel Stanciu, Daniel Results from a multicentre clinical trial. Eur J Cancer Suppl
Stoicescu, Eugen Dumitru, Corina Pietrăreanu, Daniela 2008;.6:9-15.
Bartoş, Roberta Mânzat Saplacan, Iuliana Pârvulescu, 15. Sporea I, Badea R, Popescu A, et al. Contrast-enhanced ul-
Roxana Vădan, Gabriela Smira and Liliana Tuţă trasound (CEUS) for the evaluation of focal liver lesions - a
prospective multicenter study of its usefulness in clinical
References practice. Ultraschall Med 2014;35:259-266.
16. Xie L, Guang Y, Ding H, Cai A, Huang Y. Diagnostic value
1. Karhunen PJ. Benign hepatic tumours and tumour like con- of contrast-enhanced ultrasound, computed tomography
ditions in men. J Clin Pathol 1986;39:183-188. and magnetic resonance imaging for focal liver lesions: a
2. Fukukura Y, Nakashima O, Kusaba A, Kage M, Kojiro M. meta-analysis. Ultrasound Med Biol 2011;37:854-861.
Angioarchitecture and blood circulation in focal nodular 17. Guang Y, Xie L, Ding H, Cai A, Huang Y. Diagnosis value
hyperplasia of the liver. J Hepatol 1998;29:470-475. of focal liver lesions with SonoVue(R)-enhanced ultra-
146 Roxana Şirli et al CEUS for the assessment of focal nodular hyperplasia – results of a multicentre study
sound compared with contrast-enhanced computed tomog- perplasia: diagnostic performance compared with contrast-
raphy and contrast-enhanced MRI: a meta-analysis. J Can- enhanced CT. Eur Radiol 2013;23:2546-2554.
cer Res Clin Oncol 2011;137:1595-1605. 30. Denis de Senneville B, Frulio N, Laumonier H, Salut C,
18. Friedrich-Rust M, Klopffleisch T, Nierhoff J, et al. Con- Lafitte L, Trillaud H. Liver contrast-enhanced sonography:
trast-Enhanced Ultrasound for the differentiation of benign computer-assisted differentiation between focal nodular
and malignant focal liver lesions: a meta-analysis. Liver Int hyperplasia and inflammatory hepatocellular adenoma by
2013;33:739-755. reference to microbubble transport patterns. Eur Radiol
19. Giesel FL, Delorme S, Sibbel R, Kauczor HU, Krix M. 2020;30:2995-3003.
Contrast-enhanced ultrasound for the characterization of 31. Roche V, Pigneur F, Tselikas L, et al. Differentiation of
incidental liver lesions - an economical evaluation in com- focal nodular hyperplasia from hepatocellular adenomas
parison with multi-phase computed tomography. Ultra- with low-mechanical-index contrast-enhanced sonography
schall Med 2009;30:259-268. (CEUS): effect of size on diagnostic confidence. Eur Radiol
20. Sirli R, Sporea I, Martie A, Popescu A, Danila M. Contrast 2015;25:186-195.
enhanced ultrasound in focal liver lesions--a cost efficiency 32. Grazioli L, Morana G, Kirchin MA, Schneider G. Ac-
study. Med Ultrason 2010;12:280-285. curate differentiation of focal nodular hyperplasia from
21. Broker MEE, Taimr P, de Vries M, et al. Performance of hepatic adenoma at gadobenate dimeglumine-enhanced
Contrast-Enhanced Sonography Versus MRI With a Liv- MR imaging: prospective study. Radiology 2005;236:166-
er-Specific Contrast Agent for Diagnosis of Hepatocellu- 177.
lar Adenoma and Focal Nodular Hyperplasia. AJR Am J 33. Boozari B, Soudah B, Rifai K, et al. Grading of hypervas-
Roentgenol 2020;214:81-89. cular hepatocellular carcinoma using late phase of contrast
22. Albrecht T, Blomley M, Bolondi L, et al. Guidelines for enhanced sonography - a prospective study. Dig Liver Dis
the use of contrast agents in ultrasound. January 2004. Ul- 2011;43:484-490.
traschall Med 2004;25:249-256. 34. Liu GJ, Xu HX, Lu MD, et al. Correlation between en-
23. Claudon M, Cosgrove D, Albrecht T, et al. Guidelines and hancement pattern of hepatocellular carcinoma on real-
good clinical practice recommendations for contrast en- time contrast-enhanced ultrasound and tumour cellular
hanced ultrasound (CEUS) - update 2008. Ultraschall Med differentiation on histopathology. Br J Radiol 2007;80:321-
2008;29:28-44. 330.
24. Dietrich CF, Nolsoe CP, Barr RG, et al. Guidelines and 35. Bolondi L, Gaiani S, Celli N, et al. Characterization of
Good Clinical Practice Recommendations for Contrast En- small nodules in cirrhosis by assessment of vascularity: the
hanced Ultrasound (CEUS) in the Liver – Update 2020 – problem of hypovascular hepatocellular carcinoma. Hepa-
WFUMB in Cooperation with EFSUMB, AFSUMB, tology 2005;42:27-34.
AIUM, and FLAUS. Ultraschall Med 2020;41:562-585. 36. Sherman M. Hepatocellular carcinoma: epidemiology,
25. Dietrich CF, Averkiou M, Nielsen MB, et al. How to per- surveillance, and diagnosis. Semin Liver Dis 2010;30:3-
form Contrast-Enhanced Ultrasound (CEUS). Ultrasound 16.
Int Open 2018;4:E2-E15. 37. Ferraioli G, Wong VW, Castera L, et al. Liver Ultrasound
26. Dietrich CF, Ignee A, Greis C, Cui XW, Schreiber-Dietrich Elastography: An Update to the World Federation for Ul-
DG, Hocke M. Artifacts and pitfalls in contrast-enhanced trasound in Medicine and Biology Guidelines and Recom-
ultrasound of the liver. Ultraschall Med 2014;35:108-125. mendations. Ultrasound Med Biol 2018;44:2419-2440.
27. Seitz K, Strobel D, Bernatik T, et al. Contrast-Enhanced 38. Jang HJ, Kim TK, Wilson SR. Small nodules (1-2 cm) in
Ultrasound (CEUS) for the characterization of focal liver liver cirrhosis: characterization with contrast-enhanced ul-
lesions - prospective comparison in clinical practice: CEUS trasound. Eur J Radiol 2009;72:418-424.
vs. CT (DEGUM multicenter trial). Parts of this manuscript 39. Huang Q, Pan F, Li W, et al. Differential Diagnosis of Atyp-
were presented at the Ultrasound Dreilandertreffen 2008, ical Hepatocellular Carcinoma in Contrast-Enhanced Ultra-
Davos. Ultraschall Med 2009;30:383-389. sound Using Spatio-Temporal Diagnostic Semantics. IEEE
28. Trillaud H, Bruel JM, Valette PJ, et al. Characterization of J Biomed Health Inform 2020;24:2860-2869.
focal liver lesions with SonoVue-enhanced sonography: 40. Zheng SG, Xu HX, Liu LN, et al. Parametric imaging with
international multicenter-study in comparison to CT and contrast-enhanced ultrasound: usefulness for characteriza-
MRI. World J Gastroenterol 2009;15:3748-3756. tion of dynamic effects of microvascularization for hepa-
29. Wang W, Chen LD, Lu MD, et al. Contrast-enhanced ultra- tocellular carcinoma and focal nodular hyperplasia. Clin
sound features of histologically proven focal nodular hy- Hemorheol Microcirc 2013;55:375-389.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 147-152
DOI: 10.11152/mu-2792
1Department of Imaging Nursing Science, Graduate School of Medicine, The University of Tokyo, Tokyo, 2Katsuragi
Hospital, Osaka, 32nd department of Internal medicine, Osaka Medical College, Osaka, 4Department of Nursing
Physiology, Graduate School of Nursing, Chiba University, Chiba, 5Department of Gerontological Nursing / Wound
Care Management, Graduate School of Medicine, The University of Tokyo, Tokyo, 6Global Nursing Research Center,
Graduate School of Medicine, The University of Tokyo, Tokyo, Japan.
Abstract
Aims: While bowel preparation for colonoscopy is the key to successful examination, taking laxatives and showing stools
to others causes both physical and mental distress to the patient. Thus, an alternative method to evaluation bowel preparation
is necessary. In the current study, we studied the colonic fecal retention by ultrasonography (US) and examined the US finding
which reflected completion of BP. Material and methods: The subjects were outpatients who underwent colonoscopy. This
report summarizes the ultrasonographic images of patients who underwent multiple US examinations for all five sites of the
colon just before, during, and immediately after bowel preparation. According to the standard protocol, the patients took 2
L of polyethylene glycol-ascorbic acid as a laxative, which was discontinued when the nurse visually judged the stool to be
clear. Results: Seven patients in their 50s–80s, none of whom were unable to complete a colonoscopy due to residual feces
were included in study. Following bowel preparation, the US images showed anechoic areas with haustration in four or all five
areas of the colon. Three of the seven patients received low-dose laxatives (1.1–1.2 L); all three had watery stools in three or
more colon areas and none of them were constipated at the time of taking 1 L of laxatives. Conclusions: Completion of bowel
preparation can be assessed by the observation of anechoic areas with haustration in multiple colonic sites by ultrasonography.
Keywords: ultrasonography; colonoscopy; fecal retention; laxative; bowel preparation
Fig 2. Ultrasonographic findings related to fecal retention: a) colonic fecal retention in a 69-year-old male patient. A longitudinal
ultrasonographic image showing hyperechoic areas and acoustic shadow with haustrations on the wall of the descending colon (ar-
rows); b) colonic gas in a 59-year-old male patient. A longitudinal ultrasonographic image showing multiple reflections on the wall of
the descending colon (arrow); c) watery stool in a 83-year-old male patient. A longitudinal ultrasonographic image showing anechoic
areas, indicating water retention in the descending colon (arrows).
and 3 tablets before sleep) prescribed by their doctors. an incomplete colonoscopy. The ultrasound images at the
Three patients (42.9%) had daily bowel movements, with completion of BP showed anechoic areas with haustra-
normal stool properties and amounts. On the morning of tion in four, or all five areas of the colon. Three patients
the colonoscopy, six of the seven patients (85.7%) had (Cases 1–3) received low-dose laxatives (1.1 to 1.2 L);
a bowel movement. Time from the start of 2 L polyeth- the US images of these three patients showed anecho-
ylene glycol-ascorbic acid to taking 1 L ranged 67 to 80 ic findings indicating watery stool retention in three or
minutes, with a median time of 73 minutes. Time from more colon areas. None of the three patients were consti-
the start of 2 L polyethylene glycol-ascorbic acid to the pated at the time of consuming 1 L of laxatives.
completion of bowl preparation ranged 91 to 155 min-
utes, with a median time of 121 minutes. Discussions
The change in ultrasonographic findings, total dose
of laxative and degree of bowel cleanliness are shown in This study firstly showed colonic US imaging at BP
Table II. None of the patients were determined to have before colonoscopy. The US results demonstrated that
four or all five areas of the colon showed anechoic areas
with haustration and the lack of hyperechoic area in the
stool indicated that BP had been completed. Patients who
received 1 L of laxative, had watery stools in three or more
of the colon sites and no fecal retention in any of these
sites, received a smaller dose of laxatives. Therefore, it is
possible to determine by US that BP has been completed.
The US images at the completion of BP showed an-
echoic areas with haustration in four or all five areas of
the colon, which indicated that BP was complete. As a
practical point for the future, endoscopy nurses should
be able to observe the colon with a hand-held US device
to assess BP completion without causing the patient the
mental distress of having to have their stool observed.
In the same way, that the effectiveness of hand-held
US observation and treatment of constipation has been
shown [20,21], the effectiveness of US observation by
endoscopy nurses needs to be studied in the near future.
There will also be a need to develop a US education pro-
Fig 3. Time course of ultrasonographic images at colon in a gram for nurses to observe colonic fecal retention [22].
69-year-old male patient (ID1). Longitudinal ultrasonograph- In addition, non-invasive assessment of BP with US may
ic images showing anechoic areas, indicating water retention
in the descending colon (arrows). The stool was judged to be help to determine the timing of BP completion for each
transparent by the nurse, and the patient finished bowel prepa- patient and minimize the physical distress caused by lax-
ration after taking 1.1 L of laxative. atives. Indeed, constipated patients have been reported
150
Table I. Patients characteristics, bowel preparation and defecation status.
ID Age Sex BMI Primary disease Laxative until Defecation status Bowel preparation
(kg/m2) the day before
Usual Morning of colonoscopy day TTT 1L (min) TTC (min)
1 69 M 19.3 HT, hyperlipidemia Sennoside 12 mg* Bowel movement: 3 times/week 1 soft stool 67 121
Stool: properties: normal
Stool amount: normal
Masaru Matsumoto et al
2 83 M 31.0 HT, diabetes, constipation Sennoside 12 mg* Bowel movement: 2-3 times/week 4 soft or watery stools 75 155
Stool: properties: normal or hard
Stool amount: normal
3 51 M 26.6 Anal polyp, hernia Sennoside 12 mg* Bowel movement: 7 times/week 3 soft stools 73 91
Stool: properties: normal
Stool amount: normal
4 50 F 41.7 HT, diabetes, reflux es- Sennoside 12 mg* Bowel movement: 7 times/week 1 soft stool 80 121
ophagitis Stool: properties: normal
Stool amount: normal
5 71 F 21.7 Bronchial asthma, bowel None Bowel movement: 7 times/week 3 soft stools 72 155
obstruction Stool: properties: soft
Stool amount: normal
6 59 M 27.2 Colon polyps, esophagitis Sennoside 12 mg* Bowel movement: 7 times/week 3 soft stools 68 124
Stool: properties: normal
Stool amount: normal
7 69 M 26.6 Chronic gastritis, Sennoside 12 mg* Bowel movement: 2 times/week None no data 105
constipation Stool: properties: normal
Stool amount: normal
BMI: body mass index, TTT: time to take, TTC: time to completion, tb: tablets, HT: hypertension; *: 3 tb after dinner and 3 tb before sleep
Evaluation of bowel preparation before colonoscopy by US monitoring of colonic fecal retention
Med Ultrason 2021; 23(2): 147-152 151
Table II. Change of colonic contents based on ultrasonographic findings, total dose of laxative and degree of bowel cleanliness.
ID Site Ultrasonographic findings Total dose of
Baseline When taking 1 L of laxative At the end of bowel preparation laxative
1 AC + watery stool watery stool 1.1L
TC + watery stool watery stool
DC - watery stool watery stool
SC watery stool watery stool watery stool
R watery stool watery stool watery stool
2 AC + watery stool watery stool 1.2 L
TC + watery stool watery stool
DC - watery stool watery stool
SC - watery stool watery stool
R - - watery stool
3 AC - watery stool watery stool 1.2 L
TC - watery stool watery stool
DC - watery stool watery stool
SC gas - watery stool
R + - -
4 AC + + watery stool 1.4L
TC - watery stool watery stool
DC - watery stool watery stool
SC - watery stool watery stool
R - - watery stool
5 AC watery stool watery stool watery stool 1.5 L
TC watery stool watery stool watery stool
DC + watery stool watery stool
SC gas watery stool watery stool
R - + watery stool
6 AC watery stool watery stool watery stool 1.6L
TC - watery stool watery stool
DC gas - watery stool
SC + gas watery stool
R - - watery stool
7 AC + no data watery stool 2.0 L
TC + no data watery stool
DC + no data watery stool
SC + no data watery stool
R - no data -
AC: Ascending colon, TC: Transverse colon, DC: Descending colon, SC: Sigmoid colon, R: Rectum, +: Hyperechoic finding indicating
fecal retention, -: No specific finding
to have less effective BP than non-constipated patients symptoms. It will be necessary to investigate whether
[11,18] and in these cases, observation-based, individu- evaluation of BP using US can reduce the amount of lax-
alized support for BP is required before colonoscopy. In atives taken by patients.
the cases IDs 1, 2 and 3, all received low doses of laxa- The present study has some limitations. Because this
tives (1.1–1.2 L); these three patients had watery stools study was based on the assumption that polyethylene
in three or more colonic areas and no fecal retention was glycol was taken internally, we did not consider the use
observed in any of these cases after 1 L of laxative. ID7 of sodium phosphate or sodium picosulfate with magne-
usually defecated twice a week, which was the least fre- sium citrate. Furthermore, the quality of BP was not as-
quent of the participants in this study. Although data are sessed from endoscopic images [23]. Future studies will
not available for 1 L of laxatives, patients with constipa- need to investigate the relationship between US images
tion may need to take more laxatives than those without and endoscopic images.
152 Masaru Matsumoto et al Evaluation of bowel preparation before colonoscopy by US monitoring of colonic fecal retention
Conclusions. Our results demonstrate that comple- colonoscopy compared with Polyethylene Glycol-Ascorbic
tion of BP can be assessed by observing the anechoic Acid (MoviPrep®). Turk J Gastroenterol 2018;29:67-74.
area with haustration observed in multiple colonic sites 11. Kunz L, Gillespie D. A Comparison of Bowel Preparations
for Colonoscopy in Constipated Adults. Gastroenterol Nurs
by US.
2017;40:364-372.
12. Matsumoto M, Tanaka S, Yabunaka K, et al. Ultrasono-
Acknowledgments: The authors are deeply grateful graphic evaluation of changes over time in one defecation
to the study participants, all of whom contributed greatly cycle in adults with functional constipation: A report of two
to this study. This research was supported by joint re- cases. Drug Discover Ther 2018;12:304-308.
search funding with the Department of Imaging Nursing 13. Yabunaka K, Matsumoto M, Yoshida M, et al. Assessment
Science. of rectal feces storage condition by a point-of-care pocket-
Conflict of interest: Masaru Matsumoto, Nao Tamai size ultrasound device for healthy adult subjects: A prelimi-
and Yuka Miura belong to a social collaboration depart- nary study. Drug Discov Ther 2018;12:42-46.
ment that receives funding from Fujifilm Corporation. 14. Yabunaka K, Matsuo J, Hara A, et al. Sonographic visuali-
zation of fecal loading in adults comparison with computed
References tomography. J Diagn Med Sonog 2015;31:86-92.
15. Yabunaka K, Nakagami G, Komagata K, Sanada H. Ultra-
1. Mihalko SL. Implementation of colonoscopy for mass sonographic follow-up of functional chronic constipation in
screening for colon cancer and colonic polyps: efficiency adults: A report of two cases. SAGE Open Med Case Rep
with high quality of care. Gastroenterol Clin North Am 2017;5:2050313x17694234.
2008;37:117-128. 16. Matsumoto M, Tsutaoka T, Nakagami G, et al. Deep learn-
2. Coriat R, Polin V, Oudjit A, et al. Gastric emptying evalua- ing-based classification of rectal fecal retention and analy-
tion by ultrasound prior colonoscopy: an easy tool following sis of fecal properties using ultrasound images in older
bowel preparation. World J Gastroenterol 2014;20:13591- adult patients. Jpn J Nurs Sci 2020;17:e12340.
13598. 17. Kawamura T, Nishiguchi T. Congenital Chloride Diarrhea
3. Harewood GC, Sharma VK, de Garmo P. Impact of colo- (CCD): A Case Report of CCD Suspected by Prenatal Ul-
noscopy preparation quality on detection of suspected co- trasonography and Magnetic Resonance Imaging (MRI).
lonic neoplasia. Gastrointest Endosc 2003;58:76-79. Am J Case Rep 2017;18:707-713.
4. Lebwohl B, Kastrinos F, Glick M, Rosenbaum AJ, Wang T, 18. Fan XP, Zhu Q, Zhou YJ, Ma T, Xia CX, Huang HL. Com-
Neugut AI. The impact of suboptimal bowel preparation on parative Study of Three Regimens of Bowel Preparation
adenoma miss rates and the factors associated with early re- Before Transabdominal Ultrasonography of the Colon. Ul-
peat colonoscopy. Gastrointest Endosc 2011;73:1207-1214. trasound Med Biol 2016;42:2140-2145.
5. Cheng CL, Liu NJ, Tang JH, et al. Predictors of Suboptimal 19. Neerincx M, Terhaar sive Droste JS, Mulder CJ, et al. Co-
Bowel Preparation Using 3-l of Polyethylene Glycol for lonic work-up after incomplete colonoscopy: significant
an Outpatient Colonoscopy: A Prospective Observational new findings during follow-up. Endoscopy 2010;42:730-
Study. Dig Dis Sci 2017;62:345-351. 735.
6. Radaelli F, Paggi S, Repici A, et al. Barriers against split- 20. Matsumoto M, Yabunaka K, Yoshida M, et al. Improve-
dose bowel preparation for colonoscopy. Gut 2017;66:1428- ment of Constipation Symptoms in an Older Adult Patient
1433. by Defecation Care Based on Using a Handheld Ultrasound
7. Rex DK, Imperiale TF, Latinovich DR, Bratcher LL. Im- Device in Home Care Settings: A Case Report. J Wound
pact of bowel preparation on efficiency and cost of colonos- Ostomy Continence Nurs 2020;47:75-78.
copy. Am J Gastroenterol 2002;97:1696-1700. 21. Matsumoto M, Yoshida M, Yabunaka K, et al. Safety and
8. Clark BT, Rustagi T, Laine L. What level of bowel prep efficacy of a defecation care algorithm based on ultrasono-
quality requires early repeat colonoscopy: systematic graphic bowel observation in Japanese home-care settings:
review and meta-analysis of the impact of preparation a single-case, multiple-baseline study. Geriatr Gerontol Int
quality on adenoma detection rate. Am J Gastroenterol 2020;20:187-194.
2014;109:1714-1723. 22. Matsumoto M, Yoshida M, Miura Y, et al. Feasibility of the
9. So H, Boo SJ, Seo H, et al. Patient descriptions of rectal ef- constipation point-of-care ultrasound educational program
fluents may help to predict the quality of bowel preparation in observing fecal retention in the colorectum: a descriptive
with photographic examples. Intest Res 2015;13:153-159. study. Jpn J Nurs Sci 2020. doi:10.1111/jjns.12385.
10. Khan MA, Patel KB, Nooruddin M, et al. Polyethylene Gly- 23. Kastenberg D, Bertiger G, Brogadir S. Bowel prepara-
col-3350 (Miralax®)+1.9-L sports drink (Gatorade®)+2 tion quality scales for colonoscopy. World J Gastroenterol
tablets of bisacodyl results in inferior bowel preparation for 2018;24:2833-2843.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 153-160
DOI: 10.11152/mu-3005
13rdPediatric Department, “Iuliu Hatieganu” University of Medicine and Pharmacy, 22nd Internal Medicine Department,
“Iuliu Hatieganu” University of Medicine and Pharmacy, 32nd Pediatric Department, “Iuliu Hatieganu” University of
Medicine and Pharmacy, 4Radiology Department, Clinical Emergency Hospital for Children, Cluj-Napoca, Romania
Abstract
Aim: To evaluate the value of abdominal ultrasonography (US) in the follow-up of paediatric patients with ulcerative
colitis (UC) compared to faecal calprotectin (FC) and colonoscopy. Material and method: In this retrospective study we
enrolled 30 paediatric patients previously diagnosed with UC, examined by abdominal US and colonoscopy within the same
week. FC was also determined during the same week. Disease activity was established using the paediatric ulcerative colitis
activity index (PUCAI). The global endoscopic activity was evaluated using the Mayo endoscopic subscore. Results: Endos-
copy revealed pathological findings of active disease in 27 out of 30 patients; 3 patients were in endoscopic remission. Only
18 of them had clinical active disease (PUCAI >10), [sensitivity (Se) 66.7% and specificity (Sp) 33% of PUCAI in detecting
endoscopic active disease). Twenty-three (76.7%) patients had FC >250 mcg/g, but in 2 of these cases the colonoscopy was
normal (Se 77.8% and Sp 33.3% in detecting active disease). At US examination, pathological findings (increased bowel wall
thickness, hypervascularity, lymphadenopathies, and/or mesenteric inflammatory fat) were found in 27 patients (90%), all with
endoscopic active disease (agreement US - colonoscopy, at patient level, k=1.0, p<0.001, Se 100% and Sp 100%). At seg-
ment level (totally 180 bowel segments examined by US), the overall agreement between US and colonoscopy was k=0.767,
p<0.001, Se 86.5%, Sp 90.1%. Of the 27 patients with US pathological findings in any of colonic segments, 23 had FC >250
mcg/g (85.1%). The inter-observer agreement for the US measurements had an overall ICC of 0.926 with p<0.001. Conclu-
sion: Abdominal US findings demonstrate a good to excellent concordance with endoscopic examination and are correlated
with elevated FC levels. Therefore, US appears as an accurate technique in assessing activity in patients with UC and might
replace colonoscopic evaluation for the follow-up.
Keywords: ultrasonography; colonoscopy; faecal calprotectin; IBD; ulcerative colitis
rectum with ascending progression in a continuous man- [29] in assessing the extension and activity of IBD [30],
ner with no small bowel involvement) with multiple biop- with a good concordance with the MRE findings [23,31].
sies. Typical endoscopic findings are represented by ery- In a recent systematic review in pediatric IBD, van Was-
thematous, bleeding, friable, erosive, ulcerative mucosa senaer et al [30] concluded that the increased bowel wall
and loss of vascularity [11,12]. The microscopic features thickness (BWT), bowel wall vascularization, presence
of UC are represented by active (neutrophils infiltration) of enlarged lymph nodes, alteration of wall stratification,
and chronic inflammation (modified architecture of the creeping fat, abscesses or other complications and the ab-
crypts and lymphoplasmacytic infiltrates) [11,13,14]. sence of colonic haustration were the major US criteria
The disease activity is currently assessed using the for IBD. The majority of the included studies regarded
paediatric ulcerative colitis activity index (PUCAI), co- both UC and CD patients (with a smaller number of UC),
lonoscopy with multiple biopsies and faecal calprotectin or only CD patients. Therefore, data about the usefulness
(FC). Colonoscopy with biopsies is the gold standard of US in paediatric UC are limited [32]. Moreover, moni-
for IBD diagnosis [15]. The recent ESPGHAN and Eu- toring IBD patients using US is not yet part of the routine
ropean Crohn’s and Colitis Organisation (ECCO) guide- standardized clinical practice, despite the good results of
lines recommend colonoscopic evaluation not only for published studies, in both children and adults [33-37].
the diagnosis but also for follow-up and assessment of Given these reasons, our aim was to analyse the value
therapeutic response (when major changes in treatment of US in following up UC paediatric patients, compared
are necessary, cancer surveillance, when FC is elevated, to FC and colonoscopy.
but the symptoms are not clearly disease-related, and in
patients with sustained clinical remission with high level Material and methods
of FC) [9]. The main limitations of endoscopy are the
need of general anaesthesia and the possible complica- This was a retrospective study, approved by the local
tions, such as gastrointestinal bleeding and perforation Ethics Committee. Written informed consent was signed
[16,17]. Colonoscopy is not easily accepted by children by the patient’s legal tutors before investigations, accord-
and their caregivers, especially the repeated procedures, ing to the hospital protocol.
performed just to confirm the mucosal healing. Study Population
FC represents a non-invasive marker of inflammation From the hospital database, we selected the paediatric
[18]. No ideal cut-off value of FC has been yet estab- patients previously diagnosed with UC, according to the
lished [19,20] but according to ESPGHAN revised Porto revised Porto criteria [11]. Patients were examined dur-
criteria in children and adolescents [11], a value above ing admission by abdominal US and colonoscopy, within
250 µg/g reflects more accurately the mucosal inflam- the same week, between January 2018 and January 2020.
mation, with sensitivity (Se), specificity (Sp), positive We excluded from the study the patients that did not pro-
predictive value (PPV) and negative predictive value vide a faecal sample. Patients without a legal tutor on
(NPV) of 71%, 100%, 100% and 47.1%, respectively admission were also excluded. Disease location and phe-
[21]. notype were determined using the Paris Classification of
Among the imaging techniques, Magnetic Resonance IBD [38] and data about the medications were collected.
Enterography (MRE) has high accuracy in identifying Fecal calprotectin
the bowel wall modifications and the disease complica- All patients provided a faecal sample, stored at 4 °C
tions [11,22]. However, high-quality images in children until processed [19], for the FC determination using
are more difficult to be obtained when compared to adults ELISA.
[23], due to the movement artefacts, poor acceptance of Abdominal US
the oral contrast and the need of general anaesthesia [24]. The protocol of the hospital regarding the abdomi-
The low accessibility, long examination time, need for nal US in IBD patients includes a standard preparation,
specific radiological competence and high cost are other patients being advised to fast for at least 8 hours. All the
limits of this method [25,26]. US examinations were performed by one of the two pae-
Abdominal ultrasound (US) is a non-invasive, low diatric radiologists with more than 10 years of experience
cost, accessible, reproductible and easily accepted by in bowel US, using XarioTM 200 (Cannon Medical Sys-
children imaging technique, used more and more fre- tems Corporation, Otawara, Japan) or Aplio 500 machine
quently in the recent years in assessing bowel inflamma- (Canon Medical Systems Corporation, Otawara, Japan)
tion [27,28]. It is largely used in clinical practice for ab- with linear (7-14 MHz) and convex (4-6 MHz) transduc-
dominal complaints or check-up examinations. US was ers, without oral contrast medium administration. The
found to have 79.7% sensitivity and 96.7% specificity investigation protocol included systematic evaluation
Med Ultrason 2021; 23(2): 153-160 155
Fig 1. Normal aspect of the bowel segments: a) longitudinal scan of the terminal ileum; b) transverse scan of the ascending colon;
c) longitudinal scan of the transverse colon; d) longitudinal scan of the descending colon; e) transverse aspect of the sigmoid;
f) longitudinal scan of the rectum (measurements in all cases between callipers).
of the abdomen starting with the intra-abdominal solid conclusion was reached by consensus. The disease ac-
organs using a low frequency convex probe, with the pa- tivity for each bowel segment on US was classified as
tient lying supine [27]. For a better examination of the present/absent based on the radiologist’s comprehensive
large bowel, the high frequency linear probe [39] with examination.
graded compression [40] was used, starting at the cae- Colonoscopy
cum in the right iliac fossa, after the localization of the All endoscopic evaluations were performed using
terminal ileum and progressing distally for the evaluation paediatrics OLYMPUS CV-190 (Hamburg, Germany)
of ascending, transverse, descending and sigmoid colon endoscopes. All patients underwent bowel preparation
[27]. Representative images from each bowel segment, with osmotic laxative, the afternoon before the colonos-
normal or pathological (fig 1 and 2), were stored in each copy [42]. On the day of examination, only clear liquids
case in the Picture Archiving Communication System were permitted. The colonoscopy procedure and the as-
(PACS), as the standard protocol of our hospital requires. sessment of inflammation, disease localization (rectum,
The images stored in PACS were analysed individu- sigmoid colon, descending colon, transverse, ascending
ally by the two radiologists, blinded to the result of the colon, caecum and terminal ileum), severity and com-
colonoscopy, value of FC and PUCAI. In all bowel seg- plications were documented using a standardized work-
ments the following parameters were evaluated: BWT sheet by a paediatric gastroenterologist with more than
(normal value <3 mm for all segments), bowel vascular- 10 years of experience. The global endoscopic activity
ity (using power Doppler), presence of lymphadenopa- was evaluated using the Mayo endoscopic subscore [43]
thies and mesenteric inflammatory fat, following the (0 – normal/inactive, 1 – mild, 2 – moderate and 3 – se-
published protocols [27,41]. In case of disagreement vere disease) and the result was recorded immediately
between the examiners, the images were reviewed and after the procedure.
Fig 2. Pathological aspect of the colon: a) transverse scan of the ascending colon, color Doppler ultrasound – the walls are thickened
and hypervascularized; b) transverse scan of the proximal descending colon with increase bowel wall thickness; c) longitudinal scan
of the distal descending colon with increase bowel wall thickness; d) mesenteric inflammatory lymphadenopathies. Of note the in-
creased thickness and echogenicity of the submucosal layer.
156 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
Table III. Mayo endoscopic subscore and median of the bowel wall thickness for each bowel segment
Median BWT at US p
Mayo Score 0 1 2 3
Bowel segment
Rectum 2.3 (2.3-2.45) 3.45 (3.0-4.0) 4.0 (3.1-4.75) 3.65 (3.3-4.0) 0.043
Sigmoid colon 2.2 (2.2-2.35) 3.3 (2.5-3.8) 4.0 (3.3-4.75) 3.95 (3.4-4.5) 0.031
Descending colon 2.2 (2.1-2.2) 2.65 (2.5-3.8) 3.3 (2.5-4.25) 3.6 (3.5-3.7) 0.067
Transverse colon 2.4 (2.4-2.55) 2.6 (2.4-3.1) 2.5 (2.35-3.15) 3.15 (2.1-4.2) 0.978
Ascending colon 2.6 (2.3-2.7) 2.45 (2.3-3.5) 2.3 (2.2-2.9) 2.85 (2.2-3.5) 0.693
Results are expressed as median (range); BWT: bowel wall thickness; US: ultrasound; p< 0.05: statistically significant.
disease, at patient level, but there were some differences obtained by Dillman et al [47] in children with CD. In
between the identification of the affected bowel segments this retrospective study, the interobserver agreement was
by the two aforementioned methods. The discordances best for the maximum BWT (95%CI 0.67 [0.64-0.70])
between US and colonoscopy may occur, as described and relatively weak for the length of involved segment
in the study of Christensen et al [44], taking into account (95%CI 0.41 [0.35-0.40]). In our UC patients, we found
that the mucosal healing is not in all cases uniform, prox- an excellent agreement between radiologists, concerning
imal to distal, but also in a patchy manner. Moreover, US the measurement of the BWT. The discrepancy with the
has the potential to examine all the layers of the bowel aforementioned study could be related mainly to the un-
wall and the extraintestinal features, whereas colonosco- derlying disease. While in CD the bowel is inflamed in a
py is limited to the visualisation of the intestinal mucosa. discontinuous and inhomogeneous manner (requiring the
The US is recognised as first line imaging technique choice of the same segment for measurement in order to
in assessing patients with IBD [45,46]. Generally, B- have a good interobserver agreement), in UC the disease
mode US and Doppler techniques are used and, in the is continuous and involves the entire circumference of
last years, elastography; the oral contrast is useful only the colon.
for small bowel investigation and intravenous contrast IBD cases may present different evolutive trends
has not been approved for paediatric population. The [48]; therefore, these patients must be monitored for
performance of US in IBD paediatric patients was evalu- long periods. The disease can progress to more extensive
ated in many studies, by comparing this method to co- forms, more frequently in paediatric patients compared
lonoscopy, histology or MRE. Barber el al [23] found to adults (29.2 % vs 20.2%) [48]. In our study, 5 children
in CD paediatric patients a good concordance between were diagnosed with pancolitis at onset and, during the
US and MRE findings, regarding the disease activity follow-up, 3 other patients were found to have pancolitis
and localisation. These authors demonstrated also high at colonoscopy. US correctly identified all these three
specificity (the highest for sigmoid colon) of both imag- cases. Due to the good agreement between US and co-
ing techniques in identifying the active disease and the lonoscopy findings on patient and on segment level, we
histology confirmed disease at bowel segment level. A concluded that US can be used not only for assessing the
significant learning curve was demonstrated for MRE, favourable evolution of the disease but also to identify
with improvement of the results after 10 examinations, flares and to establish the new extension of UC. This is
but not for US. This suggests that, for experienced ra- of utmost importance, as in these cases the colonoscopy
diologists, both methods can be used for follow-up. could be replaced by US, much easier accepted by chil-
The same good agreement between US and MRE was dren and/or parents.
158 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
Mucosal healing or endoscopic remission is the ideal replace the colonoscopy evaluation on follow-up, espe-
target for nowadays disease management strategy [49]. cially when PUCAI and FC suggest active disease.
Clinical remission is considered when PUCAI <10 [50],
Conflict of interest: none
but approximately 20% of children in Turner et al [9] and
over 50% of children in Sarbagili-Shabat et al studies References
[51] had endoscopic inflammation despite clinical inac- 1. Windsor JW, Kaplan GG. Evolving Epidemiology of
tive disease. This category of patients, being asympto- IBD. Curr Gastroenterol Rep 2019;21:40.
matic and probably with normal FC, cannot undergo co- 2. Ng SC, Shi HY, Hamidi N, et al. Worldwide incidence and
lonoscopy examination at every follow up visit. In these prevalence of inflammatory bowel disease in the 21st cen-
cases, US can bring important information about the con- tury: a systematic review of population-based studies. Lan-
dition of the colonic wall. cet 2018;390:2769-2778.
Kellar et al [52] developed a US activity score - the 3. Sýkora J, Pomahačová R, Kreslová M, Cvalínová D, Štych
simple paediatric ultrasound score (SPAUSS) - in order P, Schwarz J. Current global trends in the incidence of pedi-
to determine the most reliable parameter (mesenteric atric-onset inflammatory bowel disease. World J Gastroen-
terol 2018;24:2741-2763.
inflammatory fat, mesenteric lymph nodes, hyperemia/
4. Cabrera JM, Sato TT. Medical and surgical management
Doppler color flow and BWT) to predict bowel inflam- of pediatric ulcerative colitis. Clin Colon Rectal Surg
mation. Out of these 4 parameters, BWT and mesenteric 2018;31:71-79.
inflammatory fat had significant prediction value for the 5. Yu YR, Rodriguez JR. Clinical presentation of Crohn’s,
severity of the disease. In our study, all the children with ulcerative colitis, and indeterminate colitis: Symptoms, ex-
Mayo subscore of 3 presented mesenteric inflammatory traintestinal manifestations, and disease phenotypes. Semin
fat and BWT >3 mm in all analysed segments. Pediatr Surg 2017;26:349-355.
No definition/recommendation about US findings re- 6. North American Society for Pediatric Gastroenterology,
garding the remission in UC was published. However, Hepatology, and Nutrition; Colitis Foundation of America,
as US can detect almost all the bowel segments, assess Bousvaros A, et al. Differentiating ulcerative colitis from
Crohn disease in children and young adults: report of a
the inflammation, measure the BWT and has good cor-
working group of the North American Society for Pediat-
relation with colonoscopy findings, PUCAI and FC, the
ric Gastroenterology, Hepatology, and Nutrition and the
technique has to be considered for future analysis of IBD Crohn’s and Colitis Foundation of America. J Pediatr Gas-
in children. US should be regarded as a valuable imaging troenterol Nutr 2007;44:653-674.
technique for UC evaluation. 7. Aloi M, Cucchiara S. Extradigestive manifestations of IBD
The main limitations of our study are the retrospective in pediatrics. Eur Rev Med Pharmacol Sci 2009;13 Suppl
design, heterogeneity and the low number of patients. In 1:23-32.
addition, we did not compare the US findings, especially 8. Dotson JL, Hyams JS, Markowitz J, et al. Extraintestinal
the power Doppler signal with the histology results. The manifestations of pediatric inflammatory bowel disease and
lack of comparison of the US aspect with other imag- their relation to disease type and severity. J Pediatr Gastro-
ing techniques, especially, MRE, is another limit of our enterol Nutr 2010;51:140-145.
9. Turner D, Ruemmele FM, Orlanski-Meyer E, et al. Man-
study. It would have been of interest to compare US data
agement of paediatric ulcerative colitis, part 1: ambulatory
at diagnosis and at the inclusion in the study, but due to care-an evidence-based guideline from European Crohn’s
lack of information regarding the US data at the diagno- and Colitis Organization and European Society of Paediat-
sis in some patients, we could not realize this analysis. ric Gastroenterology, Hepatology and Nutrition. J Pediatr
Due to the retrospective design of our study, we did not Gastroenterol Nutr 2018;67:257-291.
analyse whether the US findings should prompt the clini- 10. Greuter T, Bertoldo F, Rechner R, et al. Extraintestinal
cian to reconsider the treatment options. manifestations of pediatric inflammatory bowel disease:
prevalence, presentation, and anti-TNF Treatment. J Pediatr
Conclusion Gastroenterol Nutr 2017;65:200-206.
11. Levine A, Koletzko S, Turner D, et al. ESPGHAN revised
porto criteria for the diagnosis of inflammatory bowel dis-
In conclusion, due to the good to excellent agreement
ease in children and adolescents. J Pediatr Gastroenterol
between abdominal US and colonoscopy findings on pa-
Nutr 2014;58:795-806.
tient and bowel segment levels, US should be considered 12. Spiceland CM, Lodhia N. Endoscopy in inflammatory
as a useful imaging technique to assess and follow-up bowel disease: Role in diagnosis, management, and treat-
UC patients. The good correlation between elevated FC ment. World J Gastroenterol 2018;24:4014-4020.
level and pathological US findings confirms the value of 13. Van Limbergen J, Russell RK, Drummond HE, et al.
US in detecting bowel inflammatory process. US might Definition of phenotypic characteristics of childhood-
Med Ultrason 2021; 23(2): 153-160 159
onset inflammatory bowel disease. Gastroenterology Recommendations of an International Panel of Experts. In-
2008;135:1114-1122. flamm Bowel Dis 2016;22:1168-1183.
14. DeRoche TC, Xiao SY, Liu X. Histological evaluation in 30. van Wassenaer EA, de Voogd FAE, van Rijn RR, et al. Di-
ulcerative colitis. Gastroenterol Rep (Oxf) 2014;2:178-192. agnostic accuracy of transabdominal ultrasound in detect-
15. Oliva S, Thomson M, de Ridder L, et al. Endoscopy in ing intestinal inflammation in paediatric IBD patients-a sys-
pediatric inflammatory bowel disease: a position paper on tematic review. J Crohns Colitis 2019;13:1501-1509.
behalf of the Porto IBD Group of the European Society for 31. Barber JL, Zambrano-Perez A, Olsen ØE, et al. Detect-
Pediatric Gastroenterology, Hepatology and Nutrition. J ing inflammation in inflammatory bowel disease - how
Pediatr Gastroenterol Nutr 2018;67:414-430. does ultrasound compare to magnetic resonance enterog-
16. Thakkar K, El-Serag HB, Mattek N, Gilger M. Complica- raphy using standardised scoring systems? Pediatr Radiol
tions of pediatric colonoscopy: a five-year multicenter ex- 2018;48:843-851.
perience. Clin Gastroenterol Hepatol 2008;6:515–520. 32. Civitelli F, Di Nardo G, Oliva S, et al. Ultrasonography
17. Navaneethan U, Kochhar G, Phull H, et al. Severe disease of the colon in pediatric ulcerative colitis: a prospective,
on endoscopy and steroid use increase the risk for bowel blind, comparative study with colonoscopy. J Pediatr
perforation during colonoscopy in inflammatory bowel dis- 2014;165:78-84.
ease patients. J Crohns Colitis 2012;6:470-475. 33. Maaser C, Petersen F, Helwig U, et al. Intestinal ultrasound
18. Pathirana WGW, Chubb SP, Gillett MJ, Vasikaran SD. Fae- for monitoring therapeutic response in patients with ul-
cal calprotectin. Clin Biochem Rev 2018;39:77-90. cerative colitis: results from the TRUST&UC study. Gut
19. Mumolo MG, Bertani L, Ceccarelli L, et al. From bench to 2020;69:1629-1636.
bedside: Fecal calprotectin in inflammatory bowel diseases 34. Bryant RV, Friedman AB, Wright EK, et al. Gastrointestinal
clinical setting. World J Gastroenterol 2018;24:3681-3694. ultrasound in inflammatory bowel disease: an underused re-
20. D’Amico F, Bonovas S, Danese S, Peyrin-Biroulet L. Re- source with potential paradigm-changing application. Gut
view article: faecal calprotectin and histologic remission in 2018;67:973-985.
ulcerative colitis. Aliment Pharmacol Ther 2020;51:689- 35. Rajagopalan A, Sathananthan D, An YK, et al. Gastrointes-
698. tinal ultrasound in inflammatory bowel disease care: Patient
21. D’Haens G, Ferrante M, Vermeire S, et al. Fecal calprotec- perceptions and impact on disease-related knowledge. JGH
tin is a surrogate marker for endoscopic lesions in inflam- Open 2019;4:267-272.
matory bowel disease. Inflamm Bowel Dis 2012;18:2218- 36. Kucharzik T, Wilkens R, Maconi G, et al. DOP10 Intestinal
2224. ultrasound response and transmural healing after 48 weeks
22. Yoon HM, Suh CH, Kim JR, et al. Diagnostic performance of treatment with Ustekinumab in Crohn’s disease: STAR-
of magnetic resonance enterography for detection of active DUST trial substudy. UEGW virtual 2020 LB12.
inflammation in children and adolescents with inflammato- 37. Kucharzik T, Helwig U, Seibold F, et al. Improvement
ry bowel disease: a systematic review and diagnostic meta- of intestinal ultrasound parameters within 12 weeks after
analysis. JAMA Pediatr 2017;171:1208-1216. treatment induction in CD and UC patients - first interim
23. Barber JL, Maclachlan J, Planche K, et al. There is good analysis of the trust beyond study. UEGW virtual 2020
agreement between MR enterography and bowel ultrasound P0430.
with regards to disease location and activity in paediatric 38. Levine A, Griffiths A, Markowitz J, et al. Pediatric modi-
inflammatory bowel disease. Clin Radiol 2017;72:590-597. fication of the Montreal classification for inflammatory
24. Mollard BJ, Smith EA, Lai ME, Phan T, Christensen RE, bowel disease: the Paris classification. Inflamm Bowel Dis
Dillman JR. MR enterography under the age of 10 years: a sin- 2011;17:1314-1321.
gle institutional experience. Pediatr Radiol 2016;46:43-49. 39. Cogley JR, O’Connor SC, Houshyar R, Al Dulaimy K.
25. Manetta R, Capretti I, Belleggia N, et al. Magnetic reso- Emergent pediatric US: what every radiologist should
nance enterography (MRE) and ultrasonography (US) in know. Radiographics 2012;32:651-665.
the study of the small bowel in Crohn’s disease: state of the 40. Puylaert JB. Mesenteric adenitis and acute terminal ile-
art and review of the literature. Acta Biomed 2019;90:38- itis: US evaluation using graded compression. Radiology
50. 1986;161:691-695.
26. Spinelli A, Allocca M, Jovani M, Danese S. Review article: 41. Spalinger J, Patriquin H, Miron MC, et al. Doppler US in pa-
optimal preparation for surgery in Crohn’s disease. Aliment tients with crohn disease: vessel density in the diseased bow-
Pharmacol Ther 2014;40:1009-1022. el reflects disease activity. Radiology 2000;217:787-791.
27. Elliott CL, Maclachlan J, Beal I. Paediatric bowel ul- 42. Adamiak T, Altaf M, Jensen MK, et al. One-day bowel
trasound in inflammatory bowel disease. Eur J Radiol preparation with polyethylene glycol 3350: an effective
2018;108:21-27. regimen for colonoscopy in children. Gastrointest Endosc
28. Kucharzik T, Kannengiesser K, Petersen F. The use of ul- 2010;71:573-577.
trasound in inflammatory bowel disease. Ann Gastroenterol 43. Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral
2017;30:135-144. 5-aminosalicylic acid therapy for mildly to moderately ac-
29. Calabrese E, Maaser C, Zorzi F, et al. Bowel ultrasonogra- tive ulcerative colitis. A randomized study. N Engl J Med
phy in the management of Crohn’s Disease. A Review with 1987;317:1625-1629.
160 Ioana Fodor et al Following up paediatric patients with ulcerative colitis: abdominal US vs. colonoscopy and faecal calprotectin
44. Christensen B, Hanauer SB, Gibson PR, Turner JR, Hart 49. Kim S, Park S, Kang Y, Koh H. Combining faecal calpro-
J, Rubin DT. Segmental histologic normalisation occurs in tectin and sigmoidoscopy can predict mucosal healing in
ulcerative colitis but does not improve clinical outcomes. J paediatric ulcerative colitis. Eur J Gastroenterol Hepatol
Crohns Colitis 2020;14:1345-1353. 2020;32:17-21.
45. Chiorean L, Schreiber-Dietrich D, Braden B, et al. Ultra- 50. Turner D, Otley AR, Mack D, et al. Development, valida-
sonographic imaging of inflammatory bowel disease in pedi- tion, and evaluation of a pediatric ulcerative colitis activity
atric patients. World J Gastroenterol 2015;21:5231-5241. index: a prospective multicenter study. Gastroenterology
46. Arndt H, Hauenstein C, Weber MA, Däbritz J, Bierwirth 2007;133:423-432.
C. Imaging of chronic inflammatory bowel diseases in 51. Sarbagili-Shabat C, Weiner D, Wardi J, Abramas L, Yaakov
childhood and adolescence: Repetitorium. Radiologe M, Levine A. Moderate to severe endoscopic inflammation
2020;60:1085-1096. is frequent after clinical remission in pediatric ulcerative
47. Dillman JR, Smith EA, Sanchez R, et al. Prospective co- colitis. J Pediatr Gastroenterol Nutr 2020. doi:10.1097/
hort study of ultrasound-ultrasound and ultrasound-MR en- MPG.0000000000003018.
terography agreement in the evaluation of pediatric small 52. Kellar A, Wilson S, Kaplan G, DeBruyn J, Tanyingoh D,
bowel Crohn disease. Pediatr Radiol 2016;46:490-497. Novak KL. The Simple Pediatric Activity Ultrasound
48. Roda G, Narula N, Pinotti R, et al. Systematic review with Score (SPAUSS) for the accurate detection of pediatric
meta-analysis: proximal disease extension in limited ulcer- inflammatory bowel disease. J Pediatr Gastroenterol Nutr
ative colitis. Aliment Pharmacol Ther 2017;45:1481-1492. 2019;69:e1-e6.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 161-167
DOI: 10.11152/mu-2827
1Department of Radiology, Korea University College of Medicine, Ansan Hospital, Gyeonggi, 2Medical Science
Research Center, Korea University College of Medicine, Seoul, Republic of Korea
Abstract
Aims: To compare the diagnostic performance of microvascular Doppler ultrasonography (MVUS) to B-mode and con-
ventional colour Doppler US (CDUS) for detecting acute pyelonephritis (APN) lesions in children. Material and methods:
An IRB-approved retrospective study was performed. From July 2018 to January 2019, 41 APN lesions in 28 children (15
boys, 13 girls; age range, 1-196 months; mean age, 53 months) who underwent 99mTc‒dimercaptosuccinic acid renal scintig-
raphy (DMSA) or contrast-enhanced computed tomography (CECT) and US including B-mode, CDUS, and MVUS were
enrolled in this study. Three paediatric radiologists independently reviewed the B-mode, CDUS and MVUS images for the
DMSA or CECT-proven APN lesions and evaluated the lesion visibility, lesion distinguishability and diagnostic confidence
between the MVUS and CDUS images. Results: A total 41 of APN lesions were verified by DMSA (41 lesions) or CECT (3
lesions) during the same hospitalization period with renal US. Among 41 APN lesions, 52.8% was visible on B-mode, 85.4%
on CDUS, and 94.3% on MVUS (p<0.001). Comparing the extent and margins of the lesions, MVUS had better results than
CDUS in 41.5% of the lesions, CDUS had better results in 6.5% and they were equal in 52% (p<0.001). The diagnostic con-
fidence of the APN lesions was higher for MVUS than CDUS in 36.6%, higher for CDUS than MVUS in 4.9%, and equal in
the remaining 58.5% (p<0.05). The interobserver agreement was fair to moderate. Conclusions: MVUS showed improved
detectability of hypoperfused areas in paediatric APN and provided higher diagnostic confidence.
Keywords: pyelonephritis; paediatrics; ultrasonography; colour Doppler; microvascular imaging
Fig 3. A 5-month-old girl with febrile UTI. (A) DMSA scan Fig 4. A 5-month-old boy with febrile UTI. (A) DMSA scan
confirmed APN in the upper pole of the right kidney (arrow). confirmed APN in the upper pole of the right kidney (arrow). B-
B-mode (B), colour Doppler (C), and microvascular Doppler mode (B), colour Doppler (C), and microvascular Doppler (D)
(D) for the upper pole lesion. All three readers interpreted an for the upper pole lesion. Two readers interpreted an APN le-
APN lesion seen on the B, C, and D. One of them reported that sion seen on C and D, and one reader interpreted an APN lesion
D showed better lesion distinguishability and more diagnostic seen on B, C, and D. All of them reported better lesion distin-
confidence than C, but two readers reported that the lesion dis- guishability on D than C. Two of them reported that D provided
tinguishability and diagnostic confidence were equal in both C more diagnostic confidence than C, and one of them reported
and D. that the diagnostic confidence as equal between C and D.
Med Ultrason 2021; 23(2): 161-167 165
Table III. Result of proportion test for diagnostic confidence sitivity for detecting renal blood flow than CDUS. There-
Comparison CDUS > MVUS CDUS = MVUS fore, PDUS showed superior detectability of APN repre-
MVUS > CDUS <0.001 0.02 sented as hypoperfused areas in the kidneys. Recently,
MVUS = CDUS <0.001 ―
a third-generation Doppler US technique, MVUS, has
emerged to depict spatially small and temporally slow
CDUS: colour Doppler US, MVUS: microvascular Doppler US.
Inequality signs mean superiority of diagnostic confidence. blood flow using low pulse repetition frequency with an
advanced clutter filter that can differentiate microvascu-
lar flow from clutter artifacts [10,19]. Despite the relative
novelty of microvascular imaging, Some studies already
have reported its strength and clinical benefits in various
organs including liver, breast, thyroid gland, ovary, tes-
tis, brain, kidney, GI tract, lymph node, extremities and
carotid plaques [10,11,20,21].
In kidneys, MVUS has shown the superb ability to
visualize more detailed branching patterns of interlo-
bar, arcuate and interlobular arteries to the renal cortex,
compared to CDUS or PDUS [10,22,23]. Additionally,
contrast-enhanced US has emerged as a promising alter-
native to DMSA with high sensitivity and specificity for
diagnosing APN, yet it still requires intravenous admin-
istration of US contrast media and not enough studies re-
ported for clinical application in children [24].
In this study, we evaluated APN lesion visibility using
MVUS technique called MV-FlowTM and compared it to
B-mode and CDUS. We found that MVUS showed bet-
ter visibility of APN lesions than B-mode and CDUS and
Fig 5. A 17-month-old boy with febrile UTI. (A) DMSA scan provided better depiction and better distinguishability of
confirmed APN in the mid-portion of the right kidney (arrow).
B-mode (B), colour Doppler (C), and microvascular Doppler minute hypoperfused areas compared to CDUS. Since
(D) for the mid-portion lesion. All three readers interpreted an the main pathophysiology of APN is hypoperfusion at
APN lesion seen on B, C, and D. Two of them reported that the level of microvessels including cortical arteriolar
D showed better lesion distinguishability than C, and one of vasoconstriction and peritubular capillary occlusion by
them reported as equal. Two of them reported that D provided inflammatory cells, imaging of perfusion by minute ves-
more diagnostic confidence than C, and one of them reported
as equal. sels using advanced techniques such as MVUS are help-
ful in the diagnosis of APN. Thus, this study revealed
Discussion that the feasibility of MVUS to detect APN lesions was
superior to that of CDUS. Further studies with MVUS
The US findings of APN are changes in renal paren- could extend its clinical value for many vascular indica-
chymal echogenicity, swelling of the renal parenchyma, tions, including renal infarction, renal cortical necrosis,
loss of renal corticomedullary differentiation, urothelial renal tumours, chronic renal disease, acute tubular necro-
thickening of the renal pelvis and ureter and an area of sis, renal transplantation, as well as acute pyelonephritis.
decreased perfusion, which may involve the kidneys US examination is recognized as an operator-de-
focally or diffusely [8,12]. As US can also provide a pendent technique. The artifacts or factors degrading
structural overview of the renal anatomy and rule out hy- the quality of paediatric US, such as motion, crying and
dronephrosis, abscess or calculus, it is worthwhile as a bowel gas could be improved according to the radiolo-
primary imaging modality for APN in children [13,14]. gists’ experience and know‒how. To minimize operator-
However, these US findings were not always well-char- dependent issues in this study, the most experienced ra-
acterized, and a number of papers have reported better diologist performed all US scans including the B-mode,
US techniques for diagnosing paediatric APN including CDUS and MVUS scans. Despite our expectations, the
B-mode, colour Doppler, power Doppler and contrast- Fleiss Kappa values were all less than 0.6, which meant
enhanced US [2,15‒18]. Power Doppler US (PDUS) has the interobserver agreement was not perfect. For exam-
been tried for the past 14 years as a method for assessing ple, the lesion visibility in MVUS showed the highest re-
more renal vascular structures and achieving better sen- sult (94.3%). However, the kappa value among the three
166 Gayoung Choi et al Microvascular Doppler ultrasound in children with acute pyelonephritis
readers was the lowest (0.091). We suggest that these ized tomography(CT) scans and its clinical outcome. Ko-
conflicting results were due to the lack of experience rean J Intern Med 1997;12:122-127.
with MVUS in two readers and the independent interpre- 6. Hwang M, Piskunowicz M, Darge K. Advanced Ul-
trasound Techniques for Pediatric Imaging. Pediatrics
tation of the three readers without consensus. The highest
2019;143:e20182609.
kappa value was seen when the readers interpreted the B-
7. Lavocat MP, Granjon D, Allard D, Gay C, Freycon MT,
mode images, which were the basic and most familiar US Dubois F. Imaging of pyelonephritis. Pediatr Radiol
images to all three radiologists. We could expect that the 1997;27:159-165.
agreement would have been higher if more of the readers 8. Craig WD, Wagner BJ, Travis MD. Pyelonephritis: radio-
were familiar with MVUS images. logic‒pathologic review. Radiographics 2008;28:255-277.
There were several limitations to our study. First, the 9. Park AY, Seo BK. Up-to-date Doppler techniques for breast
data were retrospectively obtained. Therefore, we could tumor vascularity: superb microvascular imaging and con-
only enrol a small number of patients. Second, the sec- trast‒enhanced ultrasound. Ultrasonography 2018;37:98-
ond-generation Doppler US technique, PDUS, was not 106.
obtained in this study. The advantage of MVUS would 10. Yoo J, Je BK, Choo JY. Ultrasonographic Demonstration
of the Tissue Microvasculature in Children: Microvascular
be reinforced if a comparison between PDUS and MVUS
Ultrasonography Versus Conventional Color Doppler Ul-
was presented. Third, we did not enrol the patients af- trasonography. Korean J Radiol 2020;21:146-158.
fected by UTI without APN lesion. Fourth, all US ex- 11. 11. Jiang ZZ, Huang YH, Shen HL, Liu XT. Clinical appli-
aminations in our study were performed by one paediat- cations of superb microvascular imaging in the liver, breast,
ric radiologist. We expect that further studies performed thyroid, skeletal muscle, and carotid plaques. J Ultrasound
by more radiologists aware of MVUS may show better Med 2019;38:2811-2820.
agreement between the observers. 12. Chung EM, Soderlund KA, Fagen KE. Imaging of the Pedi-
In conclusion, MVUS showed improved detectabil- atric Urinary System. Radiol Clin North Am 2017;55:337-
ity of hypoperfused areas in paediatric APN compared to 357.
CDUS and increased US diagnostic performance of APN 13. Subcommittee on Urinary Tract Infection. Steering Com-
mittee on Quality Improvement and Management. Rob-
without radiation exposure. The more sensitive diagnosis
erts KB. Urinary tract infection: clinical practice guide-
of childhood APN is feasible with the new Doppler US
line for the diagnosis and management of the initial UTI
technique, MVUS. Confident, early diagnosis of APN in febrile infants and children 2 to 24 months. Pediatrics
using MVUS may help reduce additional radiation expo- 2011;128:595-610.
sure and also enable proper treatment to prevent chronic 14. Bae HJ, Park YH, Cho JH, Jang KM. Comparison of
sequelae of the kidneys. 99mTc‒DMSA Renal Scan and Power Doppler Ultra-
sonography for the Detection of Acute Pyelonephritis and
Acknowledgments: This study was supported by the Vesicoureteral Reflux. Child Kidney Dis 2018;22:47-51.
Korea University (K1810931). 15. Dacher JN, Pfister C, Monroc M, Eurin D, LeDosseur P.
Power Doppler sonographic pattern of acute pyelonephri-
Conflict of interest: none tis in children: comparison with CT. AJR Am J Roentgenol
1996;166:1451-1455.
References 16. Sakarya ME, Arslan H, Erkoc R, Bozkurt M, Atilla MK.
The role of power Doppler ultrasonography in the diagnosis
1. Majd M, Nussbaum Blask AR, Markle BM, et al. Acute py- of acute pyelonephritis. Br J Urol 1998;81:360-363.
elonephritis: comparison of diagnosis with 99mTc‒DMSA, 17. el Hajjar M, Launay S, Hossein‒Foucher C, Foulard M,
SPECT, spiral CT, MR imaging, and power Doppler US in Robert Y. Power Doppler sonography and acute pyelone-
an experimental pig model. Radiology 2001;218:101-108. phritis in children: comparison with Tc‒DMSA scintigra-
2. Stogianni A, Nikolopoulos P, Oikonomou I, et al. Child- phy. Arch Pediatr 2002;9:21-25.
hood acute pyelonephritis: comparison of power Doppler 18. Bykov S, Chervinsky L, Smolkin V, Halevi R, Garty I. Pow-
sonography and Tc‒DMSA scintigraphy. Pediatr Radiol er Doppler sonography versus Tc‒99m DMSA scintigraphy
2007;37:685-690. for diagnosing acute pyelonephritis in children: are these
3. Vourganti S, Agarwal PK, Bodner DR, Dogra VS. Ultra- two methods comparable? Clin Nucl Med 2003;28:198-203.
sonographic evaluation of renal infections. Radiol Clin 19. Jung HK, Park AY, Ko KH, Koh J. Comparison of the diag-
North Am 2006;44:763-775. nostic performance of power Doppler ultrasound and a new
4. Das CJ, Ahmad Z, Sharma S, Gupta AK. Multimodal- microvascular Doppler ultrasound technique (AngioPlus)
ity imaging of renal inflammatory lesions. World J Radiol for differentiating benign and malignant breast masses. J
2014;6:865-873. Ultrasound Med 2018;37:2689-2698.
5. Ha SK, Seo JK, Kim SJ, et al. Acute pyelonephritis focus- 20. Ohno Y, Fujimoto T, Shibata Y. A new era in diagnostic
ing on perfusion defects on contrast enhanced computer- ultrasound, superb microvascular imaging: preliminary
Med Ultrason 2021; 23(2): 161-167 167
results in pediatric hepato-gastrointestinal disorders. Eur J ultrasound, and contrast-enhanced ultrasound in renal pa-
Pediatr Surg 2017;27:20-25. renchymal diseases. Ultrasound Q 2018;34:250-267.
21. Ayaz E, Aslan A, İnan İ, Yıkılmaz A. Evaluation of Ovar- 23. Correas JM, Anglicheau D, Joly D, Gennisson JL, Tanter
ian Vascularity in Children by Using the “Superb Micro- M, Helenon O. Ultrasound-based imaging methods of the
vascular Imaging” Ultrasound Technique in Comparison kidney-recent developments. Kiney Int 2016;90:1199-
With Conventional Doppler Ultrasound Techniques. J Ul- 1210.
trasound Med 2019;38:2751-2760. 24. Jung HJ, Choi MH, Pai KS, Kim HG. Diagnostic perfor-
22. Quaia E, Correas JM, Mehta M, Murchison JT, Gennari mance of contrast-enhanced ultrasound for acute pyelone-
AG, van Beek EJR. Gray scale ultrasound, color Doppler phritis in children. Sci Rep 2020;10:10715.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 168-175
DOI: 10.11152/mu-2732
1Department of Obstetrics and Gynaecology, Hospital Comarcal Universitario de Vinaros, Castellón, 2Department of
Obstetrics and Gynaecology, Hospital Clínico Universitario de Valencia, 3Department of Medicine. University Jaume
I, Castellón, 4Department of Obstetrics and Gynaecology, Hospital Peset, Valencia, 5Department of Obstetrics and
Gynaecology, Clínica Universidad de Navarra, University of Navarra, Pamplona, Spain
Abstract
Aims: To compare the diagnostic performance of two ultrasound-based diagnostic systems for the classification of benign
or malignant adnexal masses, the three-step strategy and the predictive logistic regression model LR2, both proposed by the
International Ovarian Tumour Analysis (IOTA) Group. Material and methods: Prospective observational study at a single
centre that included patients diagnosed with a persistent adnexal mass by transvaginal ultrasound over a period of two years.
They were evaluated by a non-expert sonographer by applying the three-step diagnostic strategy and the LR2 predictive model
to classify the masses as benign or malignant. Patients were treated surgically or followed up for at least one year, taking as
the standard reference for benignity or malignancy the histological diagnosis of the lesion or ultrasound changes suggestive of
malignancy during the follow-up period. Sensitivity, specificity, positive and negative likelihood ratios and overall accuracy
of both systems was calculated and compared. Results: One hundred patients were included, with a mean age of 50.6 years
(range 18-87). Surgery was performed on 62 (62%) patients and 38 (38%) were managed expectantly. Eighty-three (83%)
lesions were benign and 17 (17%) were malignant. The IOTA three-step strategy presented sensitivity of 94.1% (95%CI,
86.7-98.3%) and specificity 97.6% (95%CI, 94.8-99%). The LR2 logistic regression model showed sensitivity 94.1% (95%CI,
73-98.9%) and specificity 81.9% (95%CI 72.3-88.7%). Comparison of the two systems showed a statistically significant dif-
ference in specificity in favour of the three-step strategy. Conclusions: The IOTA three-step strategy, in addition to being sim-
ple to use in clinical practice, has a high diagnostic accuracy for the classification of benignity and malignancy of the adnexal
masses, overtaking that of other predictive models such as the LR2 logistic regression model.
Keywords: adnexal masses; diagnosis; ultrasound; three-step strategy; LR2
LR2 logistic regression model spontaneously before completing one year of follow-up,
Once the lesion was classified using the three-step the patient was excluded from the study. If any ultra-
strategy, it was evaluated using the LR2 model for cal- sound scan revealed morphological changes in the lesion,
culating the malignancy probability assessing six vari- the probability of malignancy was recalculated with the
ables: (1) age, (2) ascites, (3) papillary blood flow, (4) three-step strategy and the LR2 system. In the removed
maximum solid component diameter, (5) irregular inter- lesions, the standard reference was the histological diag-
nal cystic walls and (6) acoustic shadows. The result was nosis. Tumors that were histologically diagnosed as bor-
applied to the logistic regression formula y = 1/(1 + e−z), derline were considered malignant for statistical analysis.
where z = −5.3718 + 0.0354 (1) + 1.6159 (2) + 1.1768 Statistical analysis
(3) + 0.0697 (4) + 0.9586 (5) - 2.9486 (6), using IOTA For the qualitative variables, the frequency distribu-
Models® application software, available with IOS® op- tion value for each of the categories was obtained, pre-
erating system. A probability above 10% was considered senting the data in the form of absolute counts and per-
a high risk of malignancy [22]. centages. Quantitative variables were studied following
Therapeutic attitude Kolmogorov-Smirnov analysis to determine if the distri-
The therapeutic attitude consisted of surgical treat- bution of values was normal. Quantitative variables with
ment or expectant clinical and ultrasound follow-up. normal distribution were expressed as a central tendency
Patients who showed high risk of malignancy were ad- measure with the standard deviation and range. When
dressed to the referral oncology gynaecology unit. Pa- they did not follow a normal distribution, they were pre-
tients with masses considered benign symptomatic were sented as median and range.
referred to their general gynaecologist. In both cases, the Sensitivity, specificity, and positive (LR+) and nega-
therapeutic attitude was determined by their referring tive (LR-) likelihood ratio were calculated with a 95%
gynaecologist. Asymptomatic patients with a low-risk confidence interval of the three-step model in order to
lesion were offered surgical treatment by a general gy- discriminate between benign and malignant adnexal le-
naecologist or expectant management with clinical and sions. The sensitivity and specificity obtained with the
ultrasound follow-up at 3, 6 and 12 months during the application of the three-step model were compared with
first year, and annually thereafter. the McNemar test with those obtained with the logistic
The referring gynecologist decided whether to deter- regression model LR2. A p<0.05 was considered statisti-
mine serum tumor markers during the follow-up of the cally significant for all comparisons.
patients in whom a conservative attitude was chosen. In Statistical analysis was performed using the IBM
these cases, the value of these markers was not known by SPSS version 20® for Windows (SPSS Inc., Chicago, IL).
the investigating sonographer at the time of successive
follow-up ultrasound evaluations. Results
Reference standard
Lesions managed expectantly were considered be- One hundred and two patients with an adnexal mass
nign if they did not show ultrasound changes suggesting were studied. Two (1.9%) patients were excluded, one
malignancy for at least one year. If the lesion disappeared for a follow-up shorter than 12 months and the other for
Med Ultrason 2021; 23(2): 168-175 171
being pregnant during the study. Therefore, 100 patients
were included in the final analysis. The average age was
50.6 years (SD 17.4; range 18-87 years). Fifty-eight
(58%) patients were premenopausal.
Sixty-two (62%) patients were surgically treated and
the rest had follow-up for at least 12 months. Eighty-three
(83%) lesions were benign. Out of the total surgically re-
moved adnexal masses, 45 (72.6%) were benign and 17
(27.4%) malignant. The mean follow-up time was 14.2
months (SD 2.8; range 12-22 months). All patients in the
expectant management group showed lesions that were
considered benign at the end of the follow-up, showing
no ultrasound or clinical changes suggestive of malig-
nancy. Figure 1 shows a flow chart with the patient se-
lection, their management and the final diagnostic result.
Fig 1. Study flowchart showing the selection and management
The mean age of patients with malignant masses was of patients with an adnexal mass
51.6 (SD 17.7; range 22-85), whereas the mean age for
women with benign lesions was 49.2 (SD 15.9; range II summarizes the histological diagnoses of all adnexal
18-87). Benign lesions were diagnosed in 48 (57.8%) lesions removed with their diagnostic step for the three-
premenopausal and 35 (42.2%) postmenopausal patients. step strategy and the average LR2 outcome for each his-
Malignant lesions occurred in 8 (47%) premenopausal tological group.
and 9 (53%) postmenopausal women. Figure 2 shows a flowchart with the classification
The most common histological diagnoses among of benignity or malignancy of adnexal masses with the
the benign lesions removed were: 15 (33.3%) endome- three-step strategy, indicating the diagnostic results of
triomas and 12 (26.7%) serous cystadenomas. The most each of them according to the reference standard. With
common malignant tumours were: 6 (35.3%) serous the first two steps, 91 (91%) lesions could be classified
carcinomas and 3 (17.6%) mucinous carcinomas. Table by non-expert sonographers (37 (37%) with simple de-
Table II. Histology of the adnexal masses removed surgically (n=62) with number of cases correctly and erroneously diagnosed in
each step of the three-step strategy and the average LR2 outcome for each histological group.
Histology n (%) SD SR Expert LR2 (%)
Benign masses (n=45)
Endometrioma 15 (33.3) 8 5 (+1 FP)§ 1 5.1
Serous cystadenoma 12 (26.7) 4 7 1 17.4
Teratoma 7 (15.6) 3 4 0 1.3
Mucinous cystadenoma 4 (8.9) 1 (+1 FP)§ 1 1 2
Fibroma 2 (4.4) 0 0 2 24.4
Cystadenofibroma 2 (4.4) 0 1 1 8.8
Fibrothecoma 1 (2.2) 0 1 0 8.4
Hydrosalpinx 1 (2.2) 0 0 1 35.5
Tubo-ovarian abscess 1 (2.2) 0 1 0 17.8
Malignant masses (n=17)
Serous cystadenocarcinoma 6 (35.3) 1 5 0 60.8
Mucinous adenocarcinoma 3 (17.6) 1 2 0 80.6
Serous borderline tumor 2 (11.8) 0 1 (+1 FN)¥ 0 19.9
Endometrioid carcinoma 1 (5.9) 1 0 0 88.9
Clear cell carcinoma 1 (5.9) 0 0 1 62.8
Metastasis 1 (5.9) 0 1 0 65.4
Carcinosarcoma 1 (5.9) 1 0 0 96.3
Gastrointestinal stromal tumor 1 (5.9) 0 1 0 69.1
Presacral hemangiopericytoma 1 (5.9) 0 1 0 72.8
SD: Simple descriptor (First step); SR: Simple Rules (Second step); FP: False Positive; FN: False Negative; §: One further case was diag-
nosed as malignant (false positive); ¥: One further case was diagnosed as benign (false negative)
172 Juan José Hidalgo et al Comparison of IOTA three-step strategy and logistic regression model LR2
which each of the discordant cases was classified, nine In this sense, the results of LR+ and LR- also showed
were applied the Simple Rules and four the subjective as- a higher performance of the three-step strategy compared
sessment by the expert. Table III shows the characteris- to LR2 by presenting results of 39.2 and 0.06 respec-
tics of these discordant cases between the two diagnostic tively (vs 5.2 and 0.07 of LR2), which would indicate
systems. a better practical utility of the first system, allowing the
confirmation of the presence of malignancy in the adnex-
Discussion al lesions with greater certainty, since a diagnostic test
with LR+>10 and LR-<0.1 is highly relevant and useful
Our study aims to compare the diagnostic accuracy [27].
of two ultrasound-based predictive systems, the three- So far, five external validation studies of the IOTA
step strategy and the logistic regression model LR2, both three-step strategy have been published. However, none
proposed by IOTA, in a group of patients with adnexal of them has been compared with other predictive models
masses for classification as benign or malignant of those in the same group of patients [17-21]. We chose the LR2
lesions. Our results show that the three-step strategy pre- logistic regression system as a comparison model since it
sents better diagnostic performance than the LR2 model is one of the most evaluated, best-performing and most
and, therefore, the former should preferably be used in used adnexal lesion diagnostic systems in clinical prac-
the appraisal of the adnexal masses by non-expert sonog- tice [13,15,22,23]. This is why other diagnostic models
raphers. have used this system as a reference in comparative stud-
Both systems demonstrated similar sensitivity ies, with the results being similar [28] or favourable to
(94.1%) for the prediction of malignancy, which implies LR2 [23,29-31].
that their diagnostic capacity for patients with ovarian This would show that the three-step strategy has
cancer is high. Therefore, both models would have a low proven more diagnostic accuracy than one of the refer-
rate of false negatives, meaning that few patients with a ence systems in the classification of adnexal lesions, that
malignant pathology would be diagnosed as benign, an is, the LR2.
essential parameter in the diagnosis of ovarian cancer. The main support of our study relies on its perfor-
However, we observed a statistically significant differ- mance since it has been developed under conditions
ence in specificity in favour of the three-step strategy similar to the usual clinical practice. We included both
(97.6% vs 81.9%), which would mean that fewer patients surgically treated and expectantly managed lesions with
with benign lesions would be misdiagnosed with malig- clinical and ultrasound follow-up, a therapeutic attitude
nancy, avoiding unnecessary surgical interventions and prevailing in many patients with adnexal masses. In addi-
their potential complications, as well as improving the tion, patients were evaluated by non-expert sonographers
economic costs and psychological repercussions in these and in a non-referral centre for ultrasound diagnosis of
patients. ovarian cancer, also a common clinical practice for most
174 Juan José Hidalgo et al Comparison of IOTA three-step strategy and logistic regression model LR2
patients at their initial evaluation. We are aware that the 4. Earle CC, Schrag D, Neville BA, et al. Effect of surgeon
main limitation of our study was the number of patients specialty on processes of care and outcomes for ovarian
included, as a larger sample size would have strength- cancer patients. J Natl Cancer Inst 2006;98:172-180.
5. Engelen MJ, Kos HE, Willemse PH, et al. Surgery by con-
ened our results.
sultant gynecologic oncologists improves survival in pa-
Nevertheless, our results for the three-step strategy
tients with ovarian carcinoma. Cancer 2006;106:589-598.
can be compared to those published about other diagnos- 6. Vernooij F, Heintz P, Witteveen E, van der Graaf Y. The
tic models. For example, the two most commonly used outcomes of ovarian cancer treatment are better when pro-
systems such as the Risk of Malignancy Index and the vided by gynecologic oncologists and in specialized hos-
LR2 have shown in their validation studies sensitivity pitals: A systematic review. Gynecol Oncol 2007;105:801-
and specificity of 76-87% and 57-97% [32-35] and 92- 812.
94% and 75-85%, respectively [13,22,23,30]. Therefore, 7. Timmerman D. The use of mathematical models to evaluate
these results and those obtained in comparison with LR2 pelvic masses; can they beat an expert operator? Best Pract
in our study indicate a lower diagnostic performance of Res Clin Obstet Gynaecol 2004;18:91-104.
the rest of the predictive models compared to the three- 8. Valentin L, Jurkovic D, Van Calster B, et al. Adding a sin-
gle CA 125 measurement to ultrasound imaging performed
step strategy. This fact could be validated with direct
by an experienced examiner does not improve preopera-
comparison analysis with a larger sample. tive discrimination between benign and malignant adnexal
We are aware that the main limitation of our study is masses. Ultrasound Obstet Gynecol 2009;34:345-354.
the limited number of patients included. We consider that 9. Timmerman D, Schwärzler P, Collins WP, et al. Subjective
our results can be valid with the sample of 100 patients assessment of adnexal masses with the use of ultrasonogra-
analysed, although it would be worthwhile corroborating phy: an analysis of interobserver variability and experience.
them in a larger sample. Another limitation of our study Ultrasound Obstet Gynecol 1999;13:11-16.
may have been that it was carried out in a single hospital, 10. Sokalska A, Timmerman D, Testa AC, Van Holsbeke C,
in our case a primary-level hospital without a special- Lissoni AA, Leone FPG. Diagnostic accuracy of trans-
ized ovarian cancer unit or expert sonographers, so the vaginal ultrasound examination for assigning a specific
diagnosis to adnexal masses. Ultrasound Obstet Gynecol
inconclusive cases were evaluated by an expert sonogra-
2009;34:462-470.
pher by means of the electronic submission of ultrasound
11. Valentin L, Hagen B, Tingulstad S, Eik-Nes S. Compari-
images, being more appropriate for the expert to evaluate son of ‘pattern recognition’ and logistic regression models
the patients personally. Therefore, the best approach for for discrimination between benign and Malignantnt pelvic
future studies would be to include a greater number of masses: a prospective cross-validation. Ultrasound Obstet
centers at different levels of care, which would also allow Gynecol 2001;18:357-365.
the sample of patients included to be expanded. 12. Lee TS, Kim JW, Park NH, Song YS, Kang SB, Lee HP.
In conclusion, our results show that the three-step Assessing clinical performance of gynaecology residents:
strategy has high diagnostic accuracy for the classifica- sonographic evaluation of adnexal masses based on mor-
tion of benign or malignant adnexal lesions, surpassing phological scoring systems. Ultrasound Obstet Gynecol
the LR2 logistic regression model. These findings sug- 2005;26:776-779.
13. Nunes N, Yazbek J, Ambler G, Hoo W, Naftalin J, Jurkovic
gest that this strategy should be chosen in initial assess-
D. Prospective evaluation of the IOTA Logistic Regression
ments of ovarian pathology by non-expert sonographers, Model (LR2) for the diagnosis of ovarian cancer. Ultra-
as it also employs easily identifiable ultrasound param- sound Obstet Gynecol 2012;40:355-359.
eters by most gynaecologists. 14. Geomini P, Kruitwagen R, Bremer GL, Cnossen J, Mol
BW. The accuracy of risk scores in predicting ovar-
Conflict of interest: none ian Malignantncy: a systematic review. Obstet Gynecol
2009;113:384-394.
References 15. Kaijser J, Sayasneh A, Van Hoorde K, et al. Presurgical di-
agnosis using mathematical models and scoring systems: a
1. Cho KR, Shih IM. Ovarian cancer. Annu Rev Pathol systematic review and meta-analysis. Hum Reprod Update
2009;4:287-313. 2014;20:449-462.
2. Giede KC, Kieser K, Dodge J, Rosen B. Who should op- 16. Ameye L, Timmerman D, Valentin L, et al. Clinically ori-
erate on patients with ovarian cancer? An evidence-based ented three-step strategy for assessment of adnexal pathol-
review. Gynecol Oncol 2005;99:447-461. ogy. Ultrasound Obstet Gynecol 2012;40:582-591.
3. Paulsen T, Kjaerheim K, Kaern J, Tretli S, Tropé C. Im- 17. Sayasneh A, Kaijser J, Preisler J, et al. A multicenter pro-
proved short-term survival for advanced ovarian, tubal, and spective external validation of the diagnostic performance
peritoneal cancer patients operated at teaching hospitals. Int of IOTA simple descriptors and rules to characterize ovar-
J Gynecol Cancer 2006;16 Suppl 1:11-17. ian masses. Gynecol Oncol 2013;130:140-146.
Med Ultrason 2021; 23(2): 168-175 175
18. Testa A, Kaijser J, Wynats L, et al. Strategies to diagnose 27. Deeks JJ, Altman DG. Diagnostics tests 4: likelihood ratios.
ovarian cancer: new evidence from phase 3 of the multicen- BMJ 2004;17;329:168-169.
tre international IOTA study. Br J Cancer 2014;111:680-688. 28. Meys EMJ, Jeelof LS, Achten NMJ, et al. Estimating risk
19. Peces Rama A, Llanos Llanos MC, Sánchez Ferrer ML, of Malignantncy in adnexal masses: external validation of
Alcázar Zambrano JL, Martínez Mendoza A, Nieto Díaz the ADNEX model and comparison with other frequent-
A. Simple descriptors and simple rules of the Interna- ly used ultrasound models. Ultrasound Obstet Gynecol
tional Ovarian Tumor Analysis (IOTA) Group: a prospec- 2017;49:784-792.
tive study of combined use for the description of adnexal 29. Kaijser J, Van Gorp T, Van Hoorde K, et al. A com-
masses. Eur J Obstet Gynecol Reprod Biol 2015;195:7-11. parison between an ultrasound-based prediction mod-
20. Alcázar JL, Pascual MA, Graupera B, et al. External vali- el (LR2) and the Risk of Ovarian Malignancy Algo-
dation of IOTA simple descriptors and simple rules for rithm (ROMA) to assess the risk of malignancy in
classifying adnexal masses. Ultrasound Obstet Gynecol women with an adnexal mass. Gynecol Oncol 2013;129:377-
2016;48:397-402. 383.
21. Hidalgo JJ, Ros F, Aubá M, et al. Prospective external 30. Sayasneh A, Wynants L, Preisler J, et al. Multicentre ex-
validation of IOTA three-step strategy to characterize and ternal validation of IOTA prediction models and RMI by
classify adnexal masses and retrospective assessment of an operators with varied training. Br J Cancer 2003;108:2448-
alternative two-step strategy using simple rules risk. Ultra- 2454.
sound Obstet Gynecol 2019;53:693-700. 31. Terrin N, Schmid CH, Griffith JL, D’Agostino RB, Selker
22. Timmerman D, Testa AC, Bourne T, et al; International HP. External validity of predictive models: a comparison
Ovarian Tumor Analysis Group. Logistic regression model of logistic regression, classification trees, and neural net-
to distinguish between the benign and malignant adnexal works. J Clin Epidemiol 2003;56:721-729.
mass before surgery: a multicenter study by the Inter- 32. Moore RG, Jabre-Raughley M, Brown AK, et al. Compari-
national Ovarian Tumor Analysis Group. J Clin Oncol son of a novel multiple marker assay versus the Risk of
2005;23:8794-8801. Malignancy Index for the prediction of epithelial ovarian
23. Van Holsbeke C, Van Calster B, Bourne T, et al. Exter- cancer in patients with a pelvic mass. Am J Obstet Gynecol
nal validation of diagnostic models to estimate the risk 2010;203:228.e1-e6.
of Malignantncy in adnexal masses. Clin Cancer Res 33. Van Gorp T, Veldman J, Van Calster B, et al. Subjective
2012;18:815-825. assessment by ultrasound is superior to the risk of Ma-
24. Education and Practical Standards Committee, European lignantncy index (RMI) or the risk of ovarian Malignant-
Federation of Societies for Ultrasound in Medicine and Bi- ncy algorithm (ROMA) in discriminating benign from
ology. Minimum training recommendations for the practice malignant adnexal masses. Eur J Cancer 2012;48:1649-
of medical ultrasound. Ultraschall Med 2006;27:79-105. 1656.
25. Timmerman D, Valentin L, Bourne T, et al;, International 34. Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudz-
Ovarian Tumor Analysis (IOTA) Group. Terms, definitions inskas JG. A risk of Malignantncy index incorporating CA
and measurements to describe the ultrasonographic features 125, ultrasound and menopausal status for the accurate pre-
of adnexal tumours: a consensus opinion from the interna- operative diagnosis of ovarian cancer. Br J Obstet Gynaecol
tional ovarian tumour analysis (IOTA) group. Ultrasound 1990;97:922-929.
Obstet Gynecol 2000;16:500-505. 35. Bailey J, Tailor A, Naik R, et al. Risk of Malignantncy in-
26. Timmerman D, Testa AC, Bourne T, et al. Simple ultra- dex for referral of ovarian cancer cases to a tertiary cen-
sound based-rules for the diagnosis of ovarian cáncer. Ul- tre: does it identify the correct cases? Int J Gynecol Cancer
trasound Obstet Gynecol 2008;31:681-690. 2006;16 Suppl 1:30-34.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 176-180
DOI: 10.11152/mu-2789
1Department of Rheumatology and Immunology, 2Department of Hematology, the Second Affiliated Hospital of
Abstract
Aim: In clinical practice, an anechoic signal was often exhibited between the volar plate (VP) of the proximal interphalan-
geal joint (PIPJ) (PIPJVP) and the flexor digitorum tendon (FDT) on ultrasound, which suggests the presence of effusions
(PIPJVP-FDT effusions). The purpose of this study was to investigate the prevalence of PIPJVP-FDT effusions and to explore
the possible mechanism preliminarily. Material and methods: A single-center, cross sectional study in hand osteoarthritis
(HOA) patients, rheumatoid arthritis (RA) patients and healthy controls was conducted. Ultrasound examination was per-
formed by the same real-time scanner with 18-MHz linear array transducer. Bilateral interphalangeal joints (IPJs) of the
thumb, 2ed, 3rd, 4th and 5th PIPJs were examined. The PIPJVP-FDT effusions was defined as an anechoic signal between the
PIPJVP and FDT in two perpendicular ultrasound planes. Results: In total, 200 patients with HOA, 78 patients with RA and
101 healthy controls were eligible for the study. 37.6% of healthy controls and 35.0% of HOA patients showed PIPJVP-FDT
effusions, while only 11.5% of RA patients had PIPJVP-FDT effusions (p<0.001). The 2ed, 3rdand 4th PIPJs showed more
PIPJVP-FDT effusions, while the IPJs of the thumbs and 5th PIPJs showed less PIPJVP-FDT effusions (p<0.05). Furthermore,
the prevalence of PIPJVP-FDT effusions in different age groups were similar in HOA patients and healthy controls. Conclu-
sion: To the best of our knowledge, this paper is the first to demonstrate that the presence of PIPJVP-FDT effusions is a very
common phenomenon in HOA patients and healthy individuals, and may be unrelated to inflammation, degeneration and age.
Keywords: effusion; ultrasound; volar plate; proximal interphalangeal joint
Material and methods Fig 1. A) Normal longitudinal sonogram of the proximal inter-
phalangeal joint; B) and C) Both longitudinal and transverse
Patients enrollment sonogram showing an anechoic signal between the volar plate
A single-center, cross sectional study in HOA pa- of the proximal interphalangeal joint and flexor digitorum ten-
don in the same individual, which suggests the presence of ef-
tients, RA patients and healthy controls was conducted at fusions. E, effusion; FDT, flexor digitorum tendon; VP, volar
the outpatient clinic of the Department of Rheumatology plate.
and Immunology, the Second Affiliated Hospital of Soo-
chow University from June 2017 to February 2018. The hands relaxed in a neutral position on a table. Bilateral
diagnosis of HOA patients in this study met the following interphalangeal joints (IPJs) of the thumb, 2ed, 3rd, 4th and
two points: pain, swelling and (or) morning stiffness in 5th PIPJs were examined. Each joint was scanned in lon-
PIPJs, and cortical protrusion (i.e. step-up) observed in gitudinal plane and transverse plane. The PIPJVP-FDT
two perpendicular ultrasound planes at the PIPJ margin. effusions was defined as an anechoic signal between the
However, HOA patients were excluded from this study PIPJVP and FDT in two perpendicular ultrasound planes
when they were combined with other joint diseases or (fig 1).
connective tissue diseases, such as RA, psoriatic arthri- Statistical analysis
tis, gout, undifferentiated arthritis, systemic lupus erythe- All variables were analyzed based on IBM Corp.
matosus, unexplained tendinitis and tenosynovitis of the Released 2012. IBM SPSS Statistics for Windows, Ver-
hand. The RA patients with PIPJs involvement fulfilled sion 21.0. Armonk, NY: IBM Corp. Continuous variables
the classification criteria of 2010 American College of were presented as means and standard deviation (normal
Rheumatology/European League Against Rheumatism distribution) or median and range (non-normal distribu-
[8]. RA patients complicated with other joint diseases tion). Categorical variables were reported as absolute fre-
or connective tissue diseases, such as HOA, psoriatic quency and percentage from the sub/group. The normal-
arthritis, systemic lupus erythematosus, were excluded. ity was tested by the Shapior-Wilk test. Since the number
It should be emphasized that health volunteers were of joints with PIPJVP-FDT effusions in the patients and
excluded if they met any of the following: recent joint controls were non-normal distributed, their differences
pain, swelling pain, or morning stiffness in PIPJs; corti- among HOA, RA and control groups were tested by the
cal protrusion observed in two perpendicular ultrasound Kruskal-Wallis test and the subsequent pairwise compar-
planes in PIPJs; cortical defect observed in two perpen- ison. The results of the pairwise comparison were adjust-
dicular ultrasound planes in PIPJs; synovial hypertrophy ed by the Benjamini-Hochberg method. The correlation
or synovial fluid within articular cavity or flexor tendon of the PIPJVP-FDT effusions number between the right
sheath of hand on ultrasound. This study was approved and left hands in each patient was analyzed by Spearman
by the Human Ethics Review Committee of the Second rank correlation analysis and the difference was tested by
Affiliated Hospital of Soochow University and written the Wilcoxon signed-rank test. Chi-square test was used
informed consent was obtained from each patient and to compare the prevalences of PIPJVP-FDT effusions
healthy control, according to the World Medical Associa- in different joints and different groups. The significance
tion Declaration of Helsinki, revised in 2000, Edinburgh. level was set at p<0.05.
Ultrasound examination
Ultrasound examination was performed by the same Results
real-time scanner (MyLab30, Esaote, Italy) with a 8-MHz
linear array transducer by a specialized in musculoskel- In total, 200 patients with HOA, 78 patients with
etal ultrasonography who was blinded to other clinical RA and 101 healthy controls were eligible for the study.
information. All the participants were seated with their Although the average age of HOA and RA group was
178 Leixi Xue et al Effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum tendon
Table I. Characteristics of baseline demographic and PIPJVP-
FDT effusions
HOA patients RA patients Controls
(n=200) (n=78) (n=101)
Female sex, n 162 (81.0) 57 (73.1) 79 (78.2)
(%)
Age (years), 56.8 ± 12.2† 50.9 ± 43.6 ± 12.8
mean ± SD 16.3†‡
Patients with 70 (35.0) 9 (11.5)s§ 38 (37.6)
PIPJVP-FDT
effusions, n (%)
Total number of 165 25 139
PIPJVP-FDT
effusions, n
Fig 2. The distribution of effusions number in RA patients, † p<0.05 compared with controls; s p<0.001 compared with con-
HOA patients and healthy controls. HOA, hand osteoarthritis; trols; ‡ p<0.05 compared with OA patients; § p<0.001 compared
RA, rheumatoid arthritis. with OA patients; HOA, hand osteoarthritis; RA, rheumatoid ar-
thritis; PIPJVP-FDT effusions, effusions between volar plate of the
older than the healthy group (p<0.05), there were no proximal interphalangeal joint and flexor digitorum tendon.
sex differences. We found that 37.6% (38 out of 101)
of healthy controls and 35.0% (70 out of 200) of HOA Table II. The number of PIPJVP-FDT effusions in the right and
patients showed PIPJVP-FDT effusions, while only left hands
11.5% (9 out of 78) of RA patients had PIPJVP-FDT HOA Control
effusions (p<0.001). The total number of PIPJVP-FDT Right hand PIPJVP-FDT effusions, 83 (50.3) 72 (51.8)
effusions was 139 in 38 healthy controls and 165 in 70 n (%)
HOA patients, but only 25 in 9 RA patients (Table I). Left hand PIPJVP-FDT effusions, 82 (49.7) 67 (48.2)
The differences of the PIPJVP-FDT effusions number n (%)
among HOA, RA and control groups were analyzed. The Total number, n 165 139
results of pairwise comparison showed that compared
HOA, hand osteoarthritis; RA, rheumatoid arthritis; PIPJVP-FDT
with that in RA patients, the median of PIPJVP-FDT ef- effusions, effusions between volar plate of the proximal inter-
fusions number was larger in HOA patients (p=0.001) phalangeal joint and flexor digitorum tendon.
and healthy controls (p<0.001) (fig 2), but there were no
differences between HOA patients and healthy controls Table III. The number of PIPJVP-FDT effusions in proximal
(p>0.05). interphalangeal joints
As the total number of PIPJVP-FDT effusions in RA HOA Control
patients was too few, we analyzed only the characteristics
of PIPJVP-FDT effusions in HOA patients and healthy PIPJVP-FDT effusions in IPJs, 2(1.2) 2 (1.4)‡
controls. The correlation between right and left hand PIP- n (%)
JVP-FDT effusions was studied and the Spearman corre- PIPJVP-FDT effusions in 2ed PIPJs, 40 (24.2)†‡ 35 (25.2)†
lation coefficient was 0.690 (p<0.01). The difference of n (%)
the PIPJVP-FDT effusions number between right and left PIPJVP-FDT effusions in 3rd PIPJs, 69 (41.8)†‡ 45 (32.4)†‡
hands was tested, and the results showed that there were n (%)
no significant statistic differences (p>0.05) (Table II). PIPJVP-FDT effusions in 4th PIPJs, 44 (26.7)†‡ 40 (28.8)†‡
The 2ed, 3rdand 4th PIPJs showed more PIPJVP-FDT effu- n (%)
sions, while the IPJs of the thumbs and 5th PIPJs showed PIPJVP-FDT effusions in 5th PIPJs, 10 (6.1) 17 (12.2)†
less PIPJVP-FDT effusions (Table III). n (%)
We further analyzed the effect of age on the presence Total number, n 165 139
of PIPJVP-FDT effusions in HOA patients and healthy
controls. An age value approximating the average age † p<0.05 compared with IPJs in the same group; ‡ p<0.05 com-
was served as the grouping point. The prevalence of pared with 5th PIPJs in the same group. HOA, hand osteoarthri-
tis; IPJs, interphalangeal joints; PIPJs, proximal interphalangeal
PIPJVP effusions in HOA patients with age < 55 and ≥
joints; RA, rheumatoid arthritis; PIPJVP-FDT effusions, effusions
55 years were 36.6% and 33.6% (p>0.05), respectively. between volar plate of the proximal interphalangeal joint and flexor
In healthy controls, there were also no differences in the digitorum tendon.
Med Ultrason 2021; 23(2): 176-180 179
prevalence of PIPJVP effusions between age < 45years the number of PIPJVP-FDT effusions in the left and right
and ≥ 45 years (34.7% versus 40.4%, p>0.05). hands was similar; moreover, there was a correlation be-
tween the left and right hands in the presence of PIPJVP-
Discussion FDT effusions. These results suggest that the appearance
of PIPJVP-FDT effusions may not be associated with the
In our study, we investigated the characteristics of amount and duration of joint movement, which is sup-
PIPJVP-FDT effusions in HOA patients, RA patients ported by the fact that the prevalence of PIPJVP-FDT
and healthy controls by ultrasound. The prevalence of effusions in younger participants was comparable with
PIPJVP-FDT effusions in healthy controls was compa- that in older people in HOA patients and healthy controls.
rable with that in HOA patients and significantly higher A limitation of this study is the lack of homogeneity
than that in RA patients. Moreover, compared with that regarding the age in the three study groups. Ultrasound is
in RA patients, the median of the PIPJVP-FDT effusion more sensitive than conventional radiography in the di-
number was larger in HOA patients and healthy controls, agnosis of HOA [12-14]. What’s more, the average age in
but there were no differences between HOA patients and the HOA group is more than 50 years old. Therefore, it is
healthy controls (p>0.05). Furthermore, age made no dif- very difficult to find age-matched controls without PIPJ
ferences in the prevalence of PIPJVP-FDT effusions in osteophytes on ultrasound. Moreover, we have verified
HOA patients and healthy controls. Therefore, the pres- that age has no effect on the presence of PIPJVP-FDT
ence of PIPJVP-FDT effusions is a very common phe- effusions, so the impact of age differences among three
nomenon in HOA patients and healthy individuals, and groups may be negligible. Another limitation is the larger
may be unrelated to inflammation, degeneration and female group in the HOA population. HOA patients were
age. recruited through outpatient clinics, suggesting that more
Our study also showed that the prevalence of PIP- women are willing to seek help from doctors than men.
JVP-FDT effusions in 2ed, 3rd and 4th PIPJs were higher This factor contributes to the fact that women outnumber
than that in IPJs and 5th PIPJs. The IPJs have a narrower men in our study. In addition, the prevalence of HOA in
range of movement, the 5th PIPJs have the lower mo- females is indeed significantly higher than that in males.
tion intensity and the middle three fingers can produce A Framingham analysis of incidence of HOA showed
a more powerful mechanical force in grasping, holding an age-standardized prevalence of 44.2% in women and
and twisting [9,10]. The above analysis may explain 37.7% in men, respectively [15]. At age 63 years, the
why the three middle PIPJs are more likely to produce HOA prevalence was 86% in females and 67% in males,
PIPJVP-FDT effusions. Therefore, the occurrence of and the difference was statistically significant (p=0.005)
PIPJVP-FDT effusions may be related to the intensity of [14]. In addition, our study could not demonstrate the va-
joint movement. In addition, the anatomical structure of lidity and interobserver reliability of PIPJVP-FDT effu-
the thumb is distinctly different from the other fingers, so sions on ultrasound, so these will need further investiga-
the different anatomy may also contribute to the different tion in a new cohort study.
prevalence of effusions. In conclusion, to the best of our knowledge, this is
The intensity of joint movement may also interpret the first study that demonstrated that the presence of
the differences in PIPJVP-FDT effusions among differ- PIPJVP-FDT effusions is a very common phenomenon
ent groups. RA patients often suffer from severe hand in HOA patients and healthy individuals and this may
joint pain, dysfunction and even deformity, thus sig- be related to the movement intensity and the anatomical
nificantly restricting the movement of hand joints [11]; structure of joint rather than inflammation, degeneration
therefore, RA patients had the lower prevalence and me- and age. However, the exact mechanism and significance
dian number of PIPJVP-FDT effusions. In HOA patients, of PIPJVP-FDT effusions remain unknown and further
the movement intensity of hand joints decreased only investigations are required.
slightly due to pain and morning stiffness, so there were
no significant differences in the prevalence and median Conflicts of interest: none.
number of PIPJVP-FDT effusions between HOA patients Acknowledgments: This work was supported
and healthy controls. by the National Nature Science Foundation of Chi-
Although we did not study the role of the dominant na (81800622); the Natural Science Foundation of
hand in the pathogenesis of PIPJVP-FDT effusions, con- the Jiangsu Higher Education Institutions of China
sidering that more people are right-handed, there should (18KJB320020); the Suzhou Health and Key Talent Pro-
be more PIPJVP-FDT effusions in the right hand than in ject (GSWS2019011); and the Jiangsu social develop-
the left hand. Surprisingly, the present study showed that ment project (BE2019663).
180 Leixi Xue et al Effusions between the volar plate of the proximal interphalangeal joint and the flexor digitorum tendon
1Selçuk University, Medicine Faculty, Department of Radiology, Konya, 2Istanbul University, Cerrahpasa Medicine
Faculty, department of Radiology, Istanbul, 3Health Sciences University, Training and Research Hospital, Depart-
ment of Radiology, Konya, 4Karatay Community Health Center Family Medicine Konya, 5Health Sciences University,
Training and Research Hospital, Department of General Surgery, Konya, Turkey
Abstract
Aim: We compared the two-dimensional shear-wave elastography (2D-SWE) values between the testes with same side
operated inguinal hernia (IH) and the contralateral testes, as well as the testes of healthy volunteers without IH. Material and
methods: A total of 189 participants (117 unilateral [117 testes] and 8 bilateral operated IH patients [16 testes] and 64 healthy
volunteers [128 testes]), providing a total of 378 testicles, were investigated prospectively. All patients underwent B-mode
ultrasonography (US) and 2D-SWE examinations. Operation type, the period between diagnosis and operation, the period
since the operation, testes volumes, and 2D-SWE values were compared. Results: The B-mode US finding of the testes were
normal in all participants. The mean testes’ volume of same side operated IH was significantly lower comparing to contralat-
eral testes and the healthy group (p<0.001). The 2D-SWE values of the testes with same side operated IH were significantly
higher comparing to the contralateral testes and the healthy group (p<0.001). There was a statistically significant correlation
between 2D-SWE values and IH severity, as well as the duration of the hernia (p=0.001). There was no significant correlation
between the IH severity and testes volume (p=0.285). No significant difference was found between the direct and indirect IH
in terms of testicular volume and SWE values and between the duration of the hernia, the time after sugery, testicular volume
and SWE values according to operation techniques (p>0.005). Conclusions: The 2D-SWE can be used as an effective imag-
ing method to evaluate testicular stiffness with objective numerical values, to estimate the severity of histologic damage in
patients with operated IH.
Keywords: inguinal hernia; Shear Wave Elastography; testes; ultrasonography
Introduction two main subtypes of IH, direct and indirect, the last be-
ing more common [3]. IH can lead to serious complica-
Inguinal hernia (IH) is an important public health tions, including testicular damage and atrophy resulting
problem that requires surgical treatment and has an in- from the narrowing of the blood vessels in the spermatic
cidence rate of approximately 5 to 10% [1,2]. There are cord [4]. Testicular damage due to IH surgery has been
well documented being reported to range from 1 to 14%
Received 12.09.2020 Accepted 15.12.2020 [3]. The testicular blood supply may return to normal as
Med Ultrason the pressure effect of the hernia-induced spermatic cord
2021, Vol. 23, No 2, 181-187
Corresponding author: Fatih Ateş MD
is removed after the operation [5,6]. However, testicular
Selçuk University, Medicine Faculty, damage can be caused by injury to the spermatic cord and
Department of Radiology, Konya, Turkey vas deferens during surgery, postoperative infection, and
313 Ardıclı Mahallesi, Celal Bayar Cad., foreign body inflammatory responses to the mesh, scar,
Selçuklu, 42250 Konya, Turkey
Phone: +90 5468375296
and granulation tissue [7-9].
Fax: +903325121653 Progressive histopathological changes, such as the
E-mail: fatih_ates81@hotmail.com reduction of the germ cell count, delay germ cell matura-
182 Mehmet Sedat Durmaz et al Assessment of testes with 2D-SWE in patients with operated inguinal hernia
tion and atrophy can occur due to the effects of the IH A total of 189 participants and 378 testes (1:1 ratio of
and possible postoperative changes in the inguinal area left to right) were included in the study. Group A, total-
[10]. As a consequence, the testis becomes stiffer due ling 133 testes, was composed of 117 patients who had
to parenchymal fibrosis [11]. Ultrasonography (US) is undergone unilateral IH surgery (117 testes) and 8 pa-
a commonly used imaging technique to assess testicular tients who had undergone bilaterally IH surgery (16 tes-
pathologies. Testicular morphology, volume, and vas- tes). Group A was divided into three subgroups according
cularity can be evaluated with US. However, US results to the hernia repair operation technique applied: preperi-
may indicate no testicular damage even though histologi- toneal alloplasty with a posterior open inguinal approach
cal testicular damage is present [11,12]. Testicular biopsy (Kugel); open mesh hernia repair (Lichtenstein); and
is the gold standard technique for evaluating histological endoscopic totally extraperitoneal (TEP) mesh hernia
damage. However, this is no longer recommended be- repair. There were no reported intraoperative complica-
cause of the high likelihood that the procedure will lead tions in this group of patients. Group A was further sub-
to complications [11]. divided into IH types, either direct or indirect, accord-
Two-dimensional shear wave elastography (2D- ing to the surgical findings, and the level of IH location
SWE), a US imaging technique that provides quantitative (proximal [grade I], middle [grade II] and distal inguinal
measurement of tissue stiffness, provides an alternative canal, almost in scrotum [grade III]). The duration of the
technique for estimating histopathological abnormalities hernia in these patients and the period between the IH op-
that cannot be detected by conventional US [11]. The eration and the US examination were recorded. Group B
2D-SWE can evaluate the testis stiffness to estimate the was composed of 117 testes located on the opposite side
level of fibrosis resulting from damage has been reported of the site of IH surgery. Group C was composed of 128
[11,12]. Parenchymal damage in the testis could be as- testes from 64 healthy volunteers. The groups character-
sessed using a quantitative evaluation of the degree of istics are summarized in Table I.
parenchymal fibrosis caused by IH and possible postop- Patients with any of the following traits or conditions
erative changes in the inguinal area. were excluded from the study: less than 18 years of age;
In this study, we quantitatively evaluated the stiffness a history of any lower abdominal-pelvic surgery or or-
of testes parenchyma using 2D-SWE on the same side chiectomy; newly diagnosed IH; recurrent or failed IH
as the IH operation and we compared the results with repair; previously emergency exploration for complica-
normal testes. To the best of our knowledge, no previ- tions of IH (e.g., irreducible hernia, bowel obstruction,
ous studies have focused on the evaluation of posibile and strangulation); previous chemotherapy/radiotherapy
testicular damage in patients with operated IH using the treatment; previous abnormal US findings, such as focal
2D-SWE technique. lesion in the testicle or epididymis, varicocele, hydro-
cele, epididymitis, and orchitis.
Materials and methods US and 2D-SWE
All participants underwent B-mode US and 2D-SWE
This study was conducted between December 2017 examinations of the testes in the supine position. The US
and January 2020, after the approval of the local research and 2D-SWE examinations were performed with a high-
Ethics Committee. All subjects were informed about the frequency (4–14 MHz) linear array transducer (Aplio
study protocol and written informed consent was ob- 500, Canon Medical System Corporation, Tustin, CA).
tained before the procedure. Each patient was examined under the same standard cir-
Fig 1. A 52-year-old male healthy volunteer. The right testis was measured in three dimensions and testes volume was calculated
as 16.4 cm³. The B-mode US findings of the testis were in normal parenchymal structure (a). Quantitative 2D-SWE values were
measured in the transverse plane at the level of the testicular hilus by drawing manually contours of the entire testis structure on
propagation mode with a free ROI (b). The quantitative elasticity values were measured as 6.1 kPa (c) and 1.41 m/s (d).
There were no statistically significant differences be- (r=0.806 p=0.001 and r=0.736, p=0.001, respectively).
tween groups B and C in terms of testes volume and 2D- There was no significant correlation between the duration
SWE values (Table III). of hernia and testis volume (r=-0.004, p=0.965) and be-
There was no significant difference between the du- tween the time after the IH operation, 2D-SWE values for
ration of the hernia, the period passed after the IH op- both kPa and m/s (r=0.113, p=0.197 and r=0.081, p=0.351,
eration, testicular volume and SWE values according to respectively) and testes volume (r=0.014, p=0.869).
operation techniques (p>0.005). In addition, there was no
significant difference between the direct and indirect IH Discussion
with respect to age, duration of hernia, the period since
the IH operation, testicular volume and SWE values This study focuses on whether IH or surgery to repair
(p>0.005, Table IV). IH has any effect on testicular stiffness and testicular vol-
IH was located in proximal (grade I), middle (grade ume. We found that on the side of IH surgery the testes
II) and distal, almost in scrotum (grade III) in 14 (10.5%), were smaller and stiffer. The stiffness of testes was cor-
60 (54.5%) and 59 (44.4%) patients, respectively. There related with the grade of IH and duration of hernia but
was no statistically significant difference among these no significant correlation was found between testicular
three groups in terms of age, duration of hernia and the volume and grade of IH or the duration of the hernia.
period since the IH operation (p>0.005). There was a sig- Our results suggest that SWE can be used as an effec-
nificant positive correlation between the degree of her- tive imaging technique for assessing testicular damage
nia and 2D-SWE values of testes for both kPa and m/s in these patients. The significant correlation between 2D-
(r=0.643, p=0.001 and r=0.614, p=0.001, respectively). SWE values and the degree and the duration of the hernia
There was no correlation between the grade of hernia and supports the hypothesis that the degree of testicular dam-
testes volume (r =-0.093, p=0.285). age changes according to the effect of pressure and the
There was a positive correlation between the duration duration of the pressure on the spermatic cord and vas
of the hernia and 2D-SWE values for both kPa and m/s deferens.
Table II. The mean age, testes volume and 2D-SWE values of testes according to the group which we used.
Totaly Group A Group B Group C p
Age (year) 45.50±11.20 48.11±11.55 47.72±11.62 40.64±8.60 0.406
Volume (cm3) 15.07±4.21 13.60±3.83 15.71±4.58 15.97±3.80 0.001
Mean 2D-SWE kPa 7.04±1.81 8.54±2.08 6.22±1.08 6.30±0.90 0.001
value m/s 1.44±0.16 1.56±0.16 1.37±0.10 1.38±0.13 0.001
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), kilopascal (kPa) and meters/second (m/s).
Table III. The p value when the Group A, B and C were compared between each other acording to testis volume and 2D-SWE values.
A versus B A versus C A versus (B+C) B versus C
Testes volume 0.001 0.001 0.001 0.622
Mean 2D-SWE kPa 0.001 0.001 0.001 0.520
value m/s 0.001 0.001 0.001 0.607
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), kilopascal (kPa) and meters/second (m/s).
Table IV. The mean age, volume and 2D-SWE values of testes in Group A according to type of inguinal hernia and operation type
Direct IH Indirect IH p Lichtenstein TEP Kugel p
Age (year) 49.39±9.13 47.69±12.25 0.465 47.27±13.14 50.37±9.73 47.44±5.43 0.543
Duration of IH (month) 4.90±2.33 4.40±2.33 0.280 4.71±2.23 4.14±2.14 4.44±3.09 0.297
Time from surgery (month) 4.27±2.21 4.23±2.44 0.929 3.78±2.28 5.37±2.17 4.05±2.55 0.828
Testes volume (cm3) 14.24±3.51 13.39±3.92 0.268 13.61±4.01 13.41±3.86 13.95±3.06 0.893
Mean 2D-SWE kPa 9.02±2.51 8.38±11.90 0.122 8.67±2.03 8.26±1.29 8.46±3.28 0.620
value m/s 1.61±0.20 1.54±0.14 0.080 1.58±0.16 1.54±0.11 1.54±0.21 0.414
Two-dimensional shear-wave elastography (2D-SWE), inguinal hernia (IH), endoscopic totally extraperitoneal (TEP), kilopascal (kPa) and
meters/second (m/s).
Med Ultrason 2021; 23(2): 181-187 185
Since the effect of pressure is removed after the op- forms. This method of evaluation limits the repeatabil-
eration, the testicular damage can be prevented by return- ity and reduces the reliability of SWE values [11]. We
ing the testicular blood supply to normal [6,13]. Howev- used the average SWE values obtained by drawing the
er, IH repair operations can lead to complications, which whole testis parenchyma with free ROI in the transverse
may arise during surgery or during the postoperative pe- plane. We consider this measurement method superior,
riod that may cause damage to the spermatic cord and more applicable, reliable and more repeatable than other
vas deferens [14]. Moreover, a foreign body inflamma- methods used, in diagnosis and follow-up, under daily
tory response to the mesh may result in testicular damage out-patient practice. A single study evaluated the testicles
by causing fibrosis in the tissues surrounding the sper- of patients with postoperative IH using compressive elas-
matic cord and vas deferens [7-9,14]. The incidence of tography and showed that testicle stiffness was higher
testicular damage following IH surgery has been reported on the operated side due to testicular damage [24]. The
to range from 1 to 14% [3]. Thus, both IH and IH repair shortcomings of compressive elastography are its opera-
surgery can cause testicular damage [14]. tor dependency and the semi-quantitative nature of the
Assessment of damage to the testicle with serum hor- data, which may limit its reproducibility. However, 2D-
mone levels and sexual function is not reliable, especially SWE is less operator-dependent, real-time, reproducible
in cases where unilateral IH is present [12]. B-mode US and allows the quantitative evaluation of the target tissue
is the most commonly used initial imaging technique in [11,12,23]. Elastic value E (kPa) is calculated using the
the diagnosis of testicular damage, but B-mode US can- equation E =3ρ(m/s)² (ρ refers to the tissue density, with
not provide sufficient information for a reliable diagnosis approximated value in human body as 1 g/cm³ and m/s
in every case and, in particular, has low sensitivity for refers to the shear wave propagation velocity) [25]. The
evaluating testicular damage [14,15]. Testicular atrophy quantitative elasticity values both measured in kPa and
and a decrease in testicular volume is an important in- m/s revealed good diagnostic performance, but the speci-
dicator of testicular damage in patients with IH and re- fity of the standart deviation and area under the ROC
paired IH, with a reported incidence of 0.5% [16]. Some curve of the measurement in kPa were found significantly
studies have evaluated the effectiveness of IH on testicu- higher compared to the measurement in m/s [26]. Since
lar perfusion with Doppler US using a resistive index both SWE values (m/s and kPa) could be obtained with
(RI) [15,17]. These studies have shown the testis with IH our US device, we measured both in our study. Thus, we
has a significantly higher RI [18] or that there is no sig- think that our findings can be compared with future stud-
nificant difference in the RI values [14,15,19-21]. Stud- ies using m/s or kPa.
ies have also shown that there is no significant alteration Many studies have compared testicular damage aris-
of testicular perfusion after IH operation using Doppler ing from different operating techniques. In these studies,
US [21,22]. These results suggest that Doppler US and testicular damage was compared with reference to sexual
RI are not reliable to evaluate testicular damage. In our function, testicular volume and hormone levels [7,27,28].
study, B-mode US aspect of the testicles was normal in For IH repair surgery, especially with respect to surgical
all cases. However, the statistically significant higher techniques using mesh, there are studies that support [7-
testicular stiffness values obtained by 2D-SWE in those 9, 21] or do not [19,27,29] the testicular damage caused
patients who had undergone IH surgery showed testicular by a marked foreign body inflammatory and immuno-
damage. For these reasons, the efficacy of B-mode US logical response to the mesh. In our hospital, mainly Li-
in the assessment of testicular damage is limited. In our chtenstein, Kugel and TEP IH repair surgery techniques
study, the testes volumes were significantly lower on the are used. When these three surgical techniques were
side of IH surgery. This finding supports that testicular compared, no significant difference was found between
volume may be used for the evaluation of testicular dam- testicular volume and 2D-SWE values and, therefore, the
age. However, there was a lack of a significant correla- degree of testicular damage. In addition, there was no
tion between testicular volume and the grade of hernia significant difference between the testicular volume and
and duration of hernia, suggesting that taken separately, 2D-SWE values between operations on direct or indirect
the testicular volume cannot be used to assess testicular hernia types.
damage reliably. There are some limitations in our study. In this study,
The effectiveness of 2D-SWE to evaluate stiffness for we included patients who had undergone an IH operation.
estimating the degree of fibrosis as a result of testicular If we examined the testicles at regular intervals before
damage has been shown in many studies [11,12,23]. In and after the operation with SWE, we could separately
these studies, SWE measurements were made in lim- evaluate the possible effects of IH and IH operation on
ited areas with ROI in constant small circle or square testicular damage; this may be the subject of future re-
186 Mehmet Sedat Durmaz et al Assessment of testes with 2D-SWE in patients with operated inguinal hernia
search. A biopsy was not performed in any of the patients ity, reproducibility, and clinical potential. Ultrasound Q
due to potential complications. Since the patients did not 2018;34:206–212.
present with infertility complaints, laboratory tests (e.g., 12. Hattapoğlu S, Göya C, Arslan S, et al. Evaluation of postop-
erative undescended testicles using point shear wave elas-
semen analysis, hormone levels) were not requested. All
tography in children. Ultrasonics 2016;72:191-194.
examinations were performed by an experienced radiolo-
13. Bansal VK, Krishna A, Manek P, et al. A prospective rand-
gist, and this study did not evaluate the possibility of in- omized comparison of testicular functions, sexual functions
terobserver variability. and quality of life following laparoscopic totally extra-per-
itoneal (TEP) and trans-abdominal pre-peritoneal (TAPP)
In conclusion, our study suggests that the 2D-SWE inguinal hernia repairs. Surg Endosc 2017;31:1478–1486.
technique can be effectively used as an imaging method 14. Lal P, Bansal B, Sharma R, Pradhan G. Laparoscopic TEP
for evaluating testicular damage, using objective numeri- repair of inguinal hernia does not alter testicular perfusion.
cal values in patients with postoperative IH. Long-term Hernia 2016;20:429–434.
follow-up studies with a larger number of patients are 15. Aguilar-García J, Cano-González HA, Martínez-Jiménez
required to confirm the efficacy of using 2D-SWE for MA, de la Rosa-Zapata F, Sánchez-Aguilar M. Unilateral
Lichtenstein tension-free mesh hernia repair and testicular
routine evaluation in this patient group.
perfusion: a prospective control study. Hernia 2018;22:479-
482.
Conflict of interest: none 16. Akbulut G, Serteser M, Yücel A, et al. Can laparoscopic
hernia repair alter function and volume of testis? Rand-
References omized clinical trial. Surg Laparosc Endosc Percutan Tech
2003;13:377–381.
1. Ballas K, Kontoulis T, Skouras CH, et al. Unusual findings 17. Chan MS, Teoh AY, Chan KW, Tang YC, Ng EK, Leong
in inguinal hernia surgery: Report of 6 rare cases. Hip- HT. Randomized double-blinded prospective trial of fibrin
pokratia 2009;13:169-171. sealant spray versus mechanical stapling in laparoscopic
2. Beddy P, Ridgway PF, Geoghegan T, et al. Inguinal Hernia total extraperitoneal hernioplasty. Ann Surg 2014;259:432–
Repair Protects Testicular Function: A Prospective Study 437.
of Open and Laparoscopic Herniorraphy. J Am Coll Surg 18. El-Awady SE, Elkholy AA. Beneficial effect of inguinal
2006;203:17–23. hernioplasty on testicular perfusion and sexual function.
3. Kumar S, Dikshit P. Effect of bilateraly Lichtenstein hernia Hernia 2009;13:251–258.
repair on male gonadal function: A prospective study. Ind J 19. Ramadan SU, Gokharman D, Tuncbilek I, et al. Does the
Sci Res Tech 2015;3:66-71. presence of mesh have an effect on the testicular blood flow
4. Ozdamar MY, Karakus OZ. Testicular ischemia caused after surgical repair of indirect inguinal hernia? J Clin Ul-
by incarcerated inguinal hernia in infants: incidence, trasound 2009;37:78–81.
conservative treatment procedure, and follow-up. Urol J 20. Basha MAA, Saber S, El-Hamid M Abdalla AA, et al. As-
2017;14:4030-4033. sessment of the testicular vascularity after inguinal herni-
5. Bittner R, Schwarz J. Inguinal hernia repair: current surgi- otomy in children: a prospective color Doppler study. Acta
cal techniques. Langenbecks Arch Surg 2012;397:271–282. Radiol 2020;61:128-135.
6. Bulus H, Dogan M, Tas A, Agladıoglu K, Coskun A. The 21. Stula I, Druzijanic N, Srsen D, et al. Influence of inguinal
effects of Lichtenstein tension-free mesh hernia repair on hernia mesh repair on testicular flow and sperm autoim-
testicular arterial perfusion and sexual functions. Wien Klin munity. Hernia 2012;16:417–424.
Wochenschr 2013;125:96-99. 22. Nath P, Dey S, Karim T, Jain A, Katiyar VK, Patel G. Study
7. Hallen M, Westerdahl J, Nordin P, Gunnarsson U, Sandb- of testicular perfusion after Lichtenstein hernioplasty in un-
lom G. Mesh hernia repair and male infertility: A retrospec- complicated inguinal hernia. Int Surgy J 2018;5:1104-1110.
tive register study. Surgery 2012;151:94-98. 23. Erdogan H, Durmaz MS, Arslan S, et al. Shear Wave Elas-
8. Hallen M, Sandblom G, Nordin P, Gunnarsson U, Kvist U, tography Evaluation of Testes in Patients with Varico-
Westerdahl J. Male infertility after mesh hernia repair: A cele. Ultrasound Q 2020;36:64-68.
prospective study. Surgery 2011;149:179-184. 24. Zaharko VP, Nakonechnyy АY, Fedus VR. The significance
9. Kolbe T, Hollinsky C, Walter I, Joachim A, Rulicke T. In- of qualitative compressive elastography in boys with hernia
fluence of a new self-gripping hernia mesh on male fertility inguinale. Pharma Innovation 2016;5:26-28.
in a rat model. Surg Endosc 2010;24:445-461. 25. Durmaz MS, Arslan S, Özbakır B, et al. Effectiveness
10. Sucullu I, Filiz AI, Sen B, et al. The effects of inguinal her- of Shear Wave Elastography in the diagnosis of acute
nia repair on testicular function in young adults: a prospec- pancreatitis on admission. Med Ultrason 2018;20:278-
tive randomized study. Hernia 2010;14:165–169. 284.
11. Durmaz MS, Sivri M, Sekmenli T, Kocaoğlu C, Çiftci İ. 26. Youk JH, Son EJ, Park AY, Kim JA. Shear-wave elastogra-
Experience of using shear wave elastography imaging in phy for breast masses: local shear wave speed (m/sec) ver-
evaluation of undescended testes in children: Feasibil- sus Young modulus (kPa). Ultrasonography 2014;33:34-39.
Med Ultrason 2021; 23(2): 181-187 187
27. Aydede H, Erhan Y, Sakarya A, Kara E, Ilkgul O, Can M. the vascularization of testes? J Pediatr Surg 2009;44:788-
Effect of mesh and its location on testicular flow and sper- 790.
matogenesis in patients with groin hernia. Acta Chir Belg 29. Çelebi S, Yıldız A, Üçgül A, et al. Do open repair and dif-
2003;103:607-610. ferent laparoscopic techniques in pediatric inguinal hernia
28. Palabiyik FB, Cimilli T, Kayhan A, Toksoy N. Do the ma- repairs affect the vascularization of testes? J Pediatr Surg
nipulations in pediatric inguinal hernia operations affect 2012;47:1706-1710.
Original papers Med Ultrason 2021, Vol. 23, no. 2, 188-193
DOI: 10.11152/mu-2723
1Department of Pediatric Cardiology, 2Department of Obstetrics and Gynecology, 3Pediatric Cardiology, Emergency
Institute for Cardiovascular Diseases and Transplantation, “George Emil Palade” University of Medicine, Pharmacy,
Science and Technology, Târgu Mureș, Romania
Abstract
Aim: Congenital atrioventricular block (CAVB) is an immunological condition, secondary to the transfer of maternal Ig
G antibodies from seropositive mothers. Although the presence of these antibodies is high among pregnant women, the preva-
lence of this fetal pathology is low. The aim of this paper is to analyze a series of cases with intrauterine diagnosis of CAVB
and to present their follow-up protocol. Material and method: In the period between 2013-2020, five fetuses were diagnosed
and followed up in the Pediatric Cardiology Clinic. In each of the cases, assessment of the hemodynamic status was done by
calculation of the fetal cardiovascular profile score (CVPS). In the last cases the follow-up protocol was supplemented with
longitudinal speckle tracking evaluation of the ventricular function. Results: In the present series, intrauterine death occurred
in one case; in another case resumption of atrioventricular conduction was observed. Epicardial pacemaker implantation was
required in three of the patients. Conclusion: Completing the evaluation of ventricular function with the longitudinal speckle
tracking method in fetuses and newborn patients with congenital atrioventricular block may play an important role in establish-
ing therapeutic behavior.
Keywords: fetal; congenital atrioventricular block; echocardiography
gram, and it is of crucial importance to be monitored in following echocardiographic diagnostic protocol was
pregnant women with a positive serology [3]. performed: establishing the anatomy of the fetal heart
Given the very low frequency of this congenital pa- and excluding any structural abnormalities through a
thology, the controversial data currently existing in the segmental sequential analysis; establishing the relation-
literature on treatment and evolution, we consider it im- ship between atrial and ventricular activity, using M
portant to share our experience on this topic, by present- mode (fig 1) and pulsed Doppler recorded in an artery
ing the cases from our clinical experience, respectively and a vein, most commonly in the aorta and superior
their immediate and long-term evolution. vena cava (fig 2); determination of the atrial and ven-
tricular rates; hemodynamic evaluation of the fetus by
Material and methods calculating the fetal cardiovascular profile score (CVPS)
[4,5].
In the period between 2013-2020, five fetuses were CVPS was determined taking into account each of the
diagnosed and followed with CAVB in the Pediatric Car- 5 echocardiographic elements: 1. hydrops (effusion -1;
diology Clinic from the Emergency Institute of Cardio- skin edema -2), 2. venous Doppler pulsation (atrial rever-
vascular Diseases and Heart Transplant of Târgu-Mureș. sal in ductus venosus -1 fig3; umbilical vein atrial pulsa-
In 4 of the cases mothers were asymptomatic in terms tion - 2), 3. umbilical artery Doppler flow (absent end-
of autoimmune disease, but with positive serology with diastolic flow -1, reversed diastolic flow -2), 4. heart size
anti Ro/SSA antibodies, and in one case the mother was (cardiothoracic ratio 0.35 -50% -1; cardiothoracic area>
known to have Sjögren’s syndrome. In each case, the 50% -2 - fig 4), 5. cardiac function (RV / LV shortening
Fig 3. Doppler flow in the ductus venosus with atrial reversal Fig 4. Cardiothoracic ratio
190 Liliana Gozar et al Congenital complete atrioventricular block from literature to clinical approach
Results
lowed, 2 fetuses were diagnosed with congenital com- uterine evolution [12]. Studies published so far in the
plete heart block (CCHB), thus concluding that the literature do not clarify the role of steroid therapy in
prevalence of this pathology was low, up to 1.8-2% [8]. first-degree atrioventricular block. It is difficult to prove
According to a study published by Tsubai, the risk of the progression of first-degree atrioventricular block (di-
atrioventricular block is higher in the presence of an in- agnosed echocardiographically in M-mode or by pulsed
creased titre of maternal antibodies [9]. Doppler as a prolongation of the atrioventricular interval)
The occurrence of this major conduction disorder is to second-degree atrioventricular block and even more
secondary to cardiac injury resulting from the transpla- difficult to prove the role of corticosteroid therapy in this
cental passage of maternal IgG antibodies. Although the setting [13-15].
number of seropositive mothers is high, the frequency of Most fetuses diagnosed with second-degree atrioven-
complete atrioventricular block in the fetus is still low tricular block have developed complete atrioventricular
[1,10]. block [10,16] and there are extremely few cases in which
A study published by Askanase et al in 2002 high- an improvement in atrioventricular conduction has been
lights the progressive nature of the rhythm disorders achieved, once CAVB has been diagnosed. Data pub-
in fetuses of seropositive mothers. Maternal antibodies lished in the literature regarding corticosteroid therapy
cross the placenta and produce inflammatory changes in in the setting of second-degree atrioventricular block are
the fetal myocardium, followed by changes in myocytes also controversial. Some studies report the recovery of
and fibrosis. In certain situations, the process continues atrioventricular conduction after steroid therapy in the
even after the disappearance of maternal antibodies, as context of a second-degree block [12,17,18].
evidenced by the progressive postnatal character: evolv- In a study consisting of 214 cases with high-degree
ing from a minor disorder, prolonged PR interval high- congenital heart block, published by Levesque et al [19],
lighted at birth, to the appearance of complete atrioven- the role of corticosteroids in the recovery of atrioventric-
tricular block during childhood [10]. ular conduction was not proven. In the aforementioned
Prenatal diagnosis of fetal arrhythmias is difficult and study a case of complete atrioventricular block that re-
involves an indirect assessment of fetal heart activity. It covered atrioventricular conduction without treatment
is mandatory that the diagnostic protocol includes both was also reported [19]. Of the 5 cases we followed, case
M-mode and pulsed Doppler examination, due to the fact 2 had a similar evolution, with the recovery of atrioven-
that a second-degree atrioventricular block may be diffi- tricular conduction. At the age of 24 weeks, the fetus had
cult to differentiate from blocked atrial extrasystoles [3]. a complete atrioventricular block, but at gestational age
Intrauterine fetal demise was estimated at 6% of fe- of 32 weeks, a partial resumption of atrioventricular con-
tuses with this pathology. After birth, death is rare and duction was observed. At birth, Holter ECG monitoring
often due to cardiomyopathy associated with atrioven- revealed long periods of sinus rhythm alternating with
tricular block [11]. short periods of third-degree atrioventricular block, with
Regarding the therapy of this pathology, the data supra-Hisian escape rhythm. Although the newborn was
published in literature are limited and somewhat con- asymptomatic, at 4 weeks of age, pacemaker implanta-
troversial. The most common treatment options include tion was performed. Upon follow-up, currently at the
steroids and beta-stimulation, immunoglobulins and age of 6, the patient has long episodes of sinus rhythm
hydroxychloroquine. Administration of dexamethasone with short periods of pacemaker rhythm. This evolution
may improve pleuro-pericardial effusions and intra with the regression of a complete atrioventricular block
192 Liliana Gozar et al Congenital complete atrioventricular block from literature to clinical approach
is unusual and extremely rarely described in the literature pediatric patients with complete atrioventricular block is
[20]. To the best of our knowledge, this is the second case extremely important [24].
reported in the literature, in which an important recovery The limitations of this study are given primarily by
of atrioventricular conduction was obtained. the low number of cases with complete congenital atrio-
Cuneo et al draws attention to the existence of a tran- ventricular block and the fact that the speckle tracking
sitional period until the appearance of complete atrioven- evaluation method has not yet entered into clinical prac-
tricular block and emphasizes on the importance of in- tice, thus proving its importance in the functional evalu-
trauterine diagnostics and early initiation of therapy [21]. ation of the fetal heart.
The last of the presented cases showed a favourable
evolution throughout the pregnancy, with a heart rate Conclusions
of 70 bpm, without hemodynamic changes. During the
neonatal period, heart rates of 80 bpm were maintained, Because the prevalence of this pathology is very low,
displaying an aspect of complete atrioventricular block, each clinical case must be analyzed in the light of cur-
with a supra-Hisian escape rhythm. Furthermore, the as- rent data available in the literature. The effectiveness of
sessment of ventricular function through speckle track- corticosteroid therapy is uncertain. Echocardiographic
ing, performed postnatal, showed very good longitudinal follow-up and completion of the evaluation protocol with
strain parameter values, but with a slight asynchrony in longitudinal, global and segmental speckle tracking has
the contraction of the ventricular segments. an important role in the therapeutic decision.
Despite numerous controversial data regarding cor-
ticosteroid therapy, it should not be discouraged [18]. Conflict of interest: none
Moreover, Sonnesson et al recently published a pro-
spective study and concluded that steroid therapy has a References
positive effect on both the intrauterine and postnatal evo-
lution, on the rate of the escape rhythm and could help in- 1. Clancy RM, Kapur RP, Molad Y, Askanase AD, Buyon JP.
crease the age at which the pacemaker is implanted [22]. Immunohistologic evidence supports apoptosis, IgG depo-
sition, and novel macrophage/fibroblast crosstalk in the
Given the controversy surrounding intrauterine ther-
pathologic cascade leading to congenital heart block. Ar-
apy, the monitoring the fetal heart with complete AVB is thritis Rheum 2004;50:173-182.
of extreme importance. Risk factors for an unfavourable 2. Bordachar P, Zachary W, Ploux S, Labrousse L, Haissa-
evolution are: ventricular rate below 55 bpm, the appear- guerre M, Thambo JB. Pathophysiology, clinical course,
ance of endomyocardial fibroelastosis and low shortening and management of congenital complete atrioventricular
fraction. With the help of the speckle tracking method, block. Heart Rhythm 2013;10:760-766.
the ventricular function can be evaluated more faithfully, 3. Hunter LE, Simpson JM. Atrioventricular block during fe-
as well as any early evidence of a possible ventricular tal life. J Saudi Heart Assoc. 2015;27:164-178.
dysfunction. In case 4, although the heart rate and the 4. Falkensammer CB, Paul J, Huhta JC. Fetal congestive heart
fetal cardiovascular profile score (CVPS) were low, hav- failure: correlation of Tei-index and Cardiovascular-score.
J Perinat Med 2001;29:390-398.
ing a severely altered ejection fraction, global and seg-
5. Huhta JC. Guidelines for the evaluation of heart fail-
mental longitudinal strain and strain rates showed values ure in the fetus with or without hydrops. Pediatr Cardiol
close to normal, indicating a favourable prognosis [23]. 2004;25:274-286.
Furthermore, in the last case, the good ventricular func- 6. Michaëlsson M, Engle MA. Congenital complete heart
tion, with very good longitudinal global and segmental block: an international study of the natural history. Cardio-
strain values, helped us make the decision to delay the vasc Clin 1972;4:85-101.
implantation of the pacemaker. It is known that complete 7. Kertesz NJ, Fenrich AL, Friedman RA. Congenital complete
atrioventricular block leads in time to the appearance of atrioventricular block. Tex Heart Inst J. 1997;24(4):301-
dilated cardiomyopathy, thus evaluation of ventricular 307.
function through speckle tracking can bring useful infor- 8. Brucato A, Frassi M, Franceschini F, et al. Risk of con-
mation to help therapeutic decision-making. genital complete heart block in newborns of mothers with
anti-Ro/SSA antibodies detected by counter immunoelec-
None of the 5 patients that were born presented any
trophoresis: a prospective study of 100 women. Arthritis
side effects from the administered corticosteroid therapy, Rheum 2001;44:1832-1835.
and their evolution following the pacemaker implanta- 9. Tsuboi H, Sumida T, Noma H, et al. Maternal predictive
tion was favourable, with the normalization of echocardi- factors for fetal congenital heart block in pregnant moth-
ographic parameters of the left ventricle; proving that the ers positive for anti-SS-A antibodies. Mod Rheumatol
placement of epicardial electrodes on the left ventricle in 2016;26:569-575.
Med Ultrason 2021; 23(2): 188-193 193
10. Askanase AD, Friedman DM, Copel J, et al. Spectrum and history of autoantibody-associated congenital heart block:
progression of conduction abnormalities in infants born retrospective review of the research registry for neonatal
to mothers with anti-SSA/Ro-SSB/La antibodies. Lupus lupus. Arthritis Rheum 1999;42:2335-2345.
2002;11:145-151. 18. Ciardulli A, D’Antonio F, Magro-Malosso ER, et al. Ma-
11. Clowse MEB, Eudy AM, Kiernan E, et al. The prevention, ternal steroid therapy for fetuses with second-degree im-
screening and treatment of congenital heart block from neo- mune-mediated congenital atrioventricular block: a system-
natal lupus: a survey of provider practices. Rheumatology atic review and meta-analysis. Acta Obstet Gynecol Scand
(Oxford) 2018;57(suppl_5):v9-v17. 2018;97:787-794.
12. Jaeggi ET, Fouron JC, Silverman ED, Ryan G, Smallhorn 19. Levesque K, Morel N, Maltret A, et al. Description of
J, Hornberger LK. Transplacental fetal treatment improves 214 cases of autoimmune congenital heart block: Results
the outcome of prenatally diagnosed complete atrioven- of the French neonatal lupus syndrome. Autoimmun Rev
tricular block without structural heart disease. Circulation 2015;14:1154-1160.
2004;110:1542-1548. 20. Bordachar P, Zachary W, Ploux S, Labrousse L, Haissa-
13. Friedman DM, Kim MY, Copel JA, et al; PRIDE Investi- guerre M, Thambo JB. Pathophysiology, clinical course,
gators. Utility of cardiac monitoring in fetuses at risk for and management of congenital complete atrioventricular
congenital heart block: the PR Interval and Dexametha- block. Heart Rhythm 2013;10:760-766.
sone Evaluation (PRIDE) prospective study, Circulation 21. Cuneo BF, Ambrose SE, Tworetzky W. Detection and suc-
2008;117:485-493. cessful treatment of emergent anti-SSA-mediated fetal
14. Jaeggi ET, Fouron JC, Silverman ED, Ryan G, Smallhorn atrioventricular block. Am J Obstet Gynecol 2016;215:527-
J, Hornberger LK. Transplacental fetal treatment improves 528.
the outcome of prenatally diagnosed complete atrioven- 22. Sonesson SE, Ambrosi A, Wahren-Herlenius M. Benefits of
tricular block without structural heart disease. Circulation fetal echocardiographic surveillance in pregnancies at risk
2004;110:1542-1548. of congenital heart block: single-center study of 212 anti-
15. Rein AJ, Mevorach D, Perles Z, et al. Early diagnosis and Ro52-positive pregnancies. Ultrasound Obstet Gynecol
treatment of atrioventricular block in the fetus exposed to 2019;54:87-95.
maternal anti-SSA/Ro-SSB/La antibodies: a prospective, 23. DeVore GR, Cuneo B, Klas B, Satou G, Sklansky M. Com-
observational, fetal kinetocardiogram-based study. Circula- prehensive Evaluation of Fetal Cardiac Ventricular Widths
tion 2009;119:1867-1872. and Ratios Using a 24-Segment Speckle Tracking Tech-
16. Buyon JP, Hiebert R, Copel J, et al. Autoimmune-associated nique. J Ultrasound Med 2019;38:1039-1047.
congenital heart block: demographics, mortality, morbidity 24. Tissot C, Aggoun Y, Rimensberger PC, et al. Left ventricu-
and recurrence rates obtained from a national neonatal lu- lar epicardial VVI pacing for a congenital complete heart
pus registry. J Am Coll Cardiol 1998;31:1658-1666. block with severe myocardial dysfunction: shall epicar-
17. Saleeb S, Copel J, Friedman D, Buyon JP. Comparison of dial pacing wires be positioned left? Int J Cardiol 2007;
treatment with fluorinated glucocorticoids to the natural 116:e7-e9.
Review Med Ultrason 2021, Vol. 23, no. 2, 194-202
DOI: 10.11152/mu-2652
1Department of Maxillofacial Surgery and Implantology, 2Department of Radiology, 3Department of Medical Imaging
“Prof. Dr. Octavian Fodor” Regional Institute of Gastroenterology, “Iuliu Hațieganu” University of Medicine and
Pharmacy, Cluj-Napoca, Romania
Abstract
Preoperative diagnosis of parotid tumors plays a crucial role in selecting and planning the surgical treatment. Ultrasound
(US) with its modern techniques can contribute to the differential diagnosis of parotid tumors. This paper aims to achieve a
comprehensive review of the ultrasound techniques used in the differential diagnosis of parotid tumors, based on the latest
literature data. Considering that most parotid gland tumors are located in the superficial lobe, US is frequently the first imaging
technique used for the diagnosis of parotid tumors. Sonoelastography can provide additional data on the elasticity of parotid
tumors, but there is an overlap between malignant and benign parotid tumors parameters. Contrast-enhanced ultrasound adds
value to conventional ultrasound and allows a more complete characterization of parotid tumors. Many authors have reported
promising results using contrast-enhanced ultrasound in the differential diagnosis of parotid tumors. Multiparametric ultra-
sound with a careful and systematic approach usually allows an accurate differential diagnosis of parotid tumors.
Keywords: ultrasound; elastography; contrast-enhanced ultrasound (CEUS); diagnosis; parotid gland tumors
Fig 1. a) Grayscale appearance for pleomorphic adenoma: well-defined margins, lobulated contour, acoustic enhancement, subcap-
sular location in the superficial lobe without capsular effraction; b) grayscale appearance for Warthin tumor: well defined lesion with
necrotic areas; c) low-grade malignant lesion, well defined, irregular contour, calcification and perilesional round hypoechoic lymph
node (histological proven as a collision tumor of acinic cell carcinoma and salivary duct carcinoma)
Currently, cross-sectional imaging investigations, in- tional information related to vascularization of parotid
cluding modern US techniques, are routinely used and tumors [7,5,15]. On US, pleomorphic adenoma (PMA)
studied for the differential diagnosis of parotid tumors, typically presents well circumscribed margins, predomi-
with promising results published in literature. There are nantly homogeneous echotexture, acoustic enhancement,
a few papers published focused on the evaluation of US internal focal calcifications (rarely), anechoic / cystic
techniques used in the differential diagnosis of parotid degeneration areas. A Warthin tumor (WT) usually has
tumors. This paper aims to achieve an updated, objective well-defined margins, internal cystic changes, hyper-
and comprehensive review of the US techniques used in echoic internal septation, a more inhomogeneous struc-
the diagnosis of parotid tumors based on the latest litera- ture and acoustic enhancement. Small, low-grade malig-
ture data. nant tumors may appear well defined, homogeneous and
hypoechoic, while large, high-grade malignant tumors
Diagnostic value of ultrasound techniques (US) present irregular, poorly defined margins, heterogeneous
echotexture; extra-glandular spread and pathologic cervi-
Current US techniques can contribute to the differ- cal adenopathy can be identified (fig 1, fig 2) [4,8,16,17].
ential diagnosis of parotid lesions. In 80-90% of cases On Doppler US, WTs and malignancies have a rich
tumors are localized in the superficial lobe [6-8]. Usually capsular and/or internal vascularization. PMA shows re-
US is the initial method of choice for parotid tumors di- duced vascularization [4,8,16,17].
agnosis, due to a large number of advantages (low-cost, Fischer et al in 2010 describes a predominant distri-
accessible, non-invasive and non-irradiating technique). bution of the vessels in the periphery rather than in the
The diagnostic accuracy is very high for the differen- center for PMA, compared to the WT, in which the over-
tiation between tumors and other glandular pathologies all tumor vascularization is much higher, without signifi-
such as cystic, infectious or inflammatory lesions. US
can make the differential diagnosis between benign and
malignant tumors, can locate the intra- or extra-glandular
lesions and is very helpful for the guidance of fine nee-
dle aspiration biopsy (FNAB). The technique presents a
series of limitations such as: dependence of the opera-
tor, lack of standardized sections, low specificity in as-
sessment of parotid tumors histological type, difficulty
in evaluating tumors located in the deep parotid lobe or
posterior to bone structures [1,5,7-14].
Fig 3. Doppler US: a) pleomorphic adenoma with reduced vascularization; b) highly vascularized tumor (histological proven – a
Warthin tumor) with capsular and internal vascularization; c) anarchic distribution of vessels in the case of a malignant lesion (his-
tological proven – salivary duct carcinoma).
cant differences between the central and the peripheral on B-mode and Doppler US is difficult and can lead to
part of the tumor (p<0.05) (fig 3) [16]. misdiagnosis. Zhang et al (2018) have showed that the di-
Rong et al (2014) was able to differentiate PMA from agnostic accuracy (96.1 vs 82.4%) of DWI-MRI was in-
WT based on B-mode Doppler US characteristics [13]. creased when ADC values were used in conjunction with
Rzepakowska et al (2017) managed to differentiate PMA US Doppler flow imaging features to differentiate parotid
from WT [sensitivity (Se) 61.5%, specificity (Sp) 81.5%, tumors compared to single usage of ADC values [24].
accuracy 73.1%] as well as malignant tumors from be- Conventional US is a non-invasive examination with
nign lesions (Se 60%, Sp 95.2%, accuracy 90.3%) using a facile and quick access for the clinicians, often being
B-mode and Doppler US [7]. the first recommended for patients with parotid lesions.
Gerwel et al (2015) and Kovačević (2010) et al US completes the clinical exam, allowing in the majority
showed that an ill-defined margin is the most representa- of the cases a general evaluation of the parotid tumor and
tive US parameter for malignant tumors (Sp 86%, Se an overview in establishing the clinical final diagnosis. In
37.5%) while acoustic enhancement is a benign charac- addition, it has high Se and Sp for the differentiation from
teristic; simultaneous evaluation of all US parameters other non-neoplastic lesions of the parotid glands. US al-
can increase the method sensitivity [18,19]. lows the characterization of malignant and benign tumors
In the literature a diagnostic accuracy between 61.8- based on specific US criteria for each tumor entity, but
64.7%, a Se between 38.9-88% and a Sp between 67.4- has a low Se (38.9-88%) in the differential diagnosis of
90.1% were reported using only the conventional US pa- parotid tumors, due to an overlap between benign and
rameters in differentiating malignant from benign parotid low-grade malignant tumors, according to most literature
tumors [18,20,21]. data [5,8,15,16,18,20,21,25-28]. Mansour et al (2012)
Several authors propose that in cases of benign le- suggested in their study that vascular pattern and acous-
sions confined to the superficial parotid lobe suggested tic enhancement are the most significant parameters in
by US and confirmed by the cytological result of FNAB, differentiating PMA from WT, when using the US [29].
no further investigations are required [4,9,10,14]. In addition to conventional US, modern techniques
Cheng et al (2019) showed that the US assessment of are currently available and allow the assessment of elas-
the minimum distance between the parotid fascia and the ticity or other biomechanical tissue properties. These
parotid tumor (MFTD) with a cutoff value of 2.4 mm can techniques are reunited under the name of multiparamet-
be used to differentiate the location of benign tumors in ric ultrasonography (MPUS), adding value to the con-
the parotid gland; distance of less than 2.4 mm correlates ventional US and completing some of the disadvantages
with the tumor localization in the superficial lobe of the of the method. MPUS includes the ultrasound elastogra-
parotid gland [22]. On the other hand, for a more precise phy (USE) with the two techniques- strain elastography
and objective localization of tumors in the parotid glands (SE) and shear wave elastography (SWE) as well as the
with the help of the US, a schematic diagram based on contrast-enhanced ultrasound (CEUS) [5,8,10,20,30-32].
marking the tumor location in relation to the external ear
structures was proposed. The use of this diagram helps to Ultrasound elastography
standardize and organize the US preoperative report [23].
There is still a considerable overlap between be- Ultrasound elastography (USE) is a non-invasive
nign and low-grade malignancies [8]. Therefore, differ- technique that allows a quantitative or qualitative evalu-
entiating between malignant and benign tumors based ation of the soft tissue elasticity.
Med Ultrason 2021; 23(2): 194-202 197
The use of USE in the differential diagnosis of parot- Mansour et al (2015) demonstrated that PMA pre-
id tumors is controversial. At present, the effectiveness of sents significantly increased stiffness compared to mono-
USE, as the only examination technique, is not proven in morphic adenoma, malignant tumors or cystic lesions;
the differential diagnosis of parotid tumors. Many studies however, this criterion has not been shown to be effective
have shown that malignant tumors have higher stiffness in differentiating between malignant and benign tumors
values than benign ones and PMAs are in generally stiff- [32].
er than WTs, but there is a large overlap between these On the other hand, many authors obtained promising
findings [10,11,26,28,33]. results regarding the use of USE in the differential diag-
USE brings new information on the characterization nosis of parotid tumors.
of parotid tumors, but nevertheless it has been shown that Mansour et al in 2012, using SWE with the mean
adding it to conventional US does not improve signifi- ARFI velocity values, obtained a statistically significant
cantly the differential diagnosis between parotid tumors difference (p<0.003) between PMA and WT (2.23±0.53
[29]. Initial results looking purely at USE techniques in m/s vs 2.58±0.77 m/s) [29].
the parotid glands diagnosis have been disappointing. According to Bhatia et al (2012) the mean shear mod-
Larger prospective studies are needed to determine the ulus of benign lesions was 18.3 kPa which overlapped
value of this investigation. USE recorded a pooled Se of considerably over malignant lesions (13.5 kPa). Howev-
67%, a Sp of 64% and a diagnosis odds ratio of 8.00 in er, PMAs shown to be significantly stiffer compared with
a meta-analysis [34] that evaluated the value of this in- WTs (22.5 kPa versus 16.9 kPa) [28] (fig 5).
vestigation in the differential diagnosis of parotid tumors Wierzbicka et al (2013) used USE to differentiate be-
[9,26,34,35,36]. nign from malignant parotid tumors. Malignant tumors
Dumitriu et al (2010) could not identify in their study had a mean stiffness of 146.6 kPa as opposed to benign
a sonoelastographic pattern characteristic for the PMA tumors that had a mean stiffness of 88.7 kPa. The results
[37]. Although the same authors found a statistically were statistically significant (p <0.001) and the authors
significant difference between the elastography scores concluded that adding USE to US improves the differen-
of malignant tumors compared to benign lesions, the de- tial diagnosis of parotid tumors [39].
tailed analysis was not possible to establish a cutoff point Cortcu et al evaluated the USE in discrimination of
for this scores, more likely due to the high stiffness of malignant and benign parotid tumors based on strain ra-
PMAs (fig 4) [38]. tio (SR). Median SR value for benign tumors was 1.11
Fig 4. Strain elastography technique showing increased stiffness in both pleomorphic adenoma (a) and a malignant tumor (b)
Fig 5. Shear-wave elastography. Illustration of increased stiffness of pleomorphic adenoma (a) compared with Warthin tumor (b)
198 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
and 2.75 for malignant tumors, with a statistically signifi- tumors was 2.75±0.95, compared with malignant tumors
cant p value. For the SR cutoff value of 2.1, the Se was score of 3.44±0.85; (p<0.034) [43]. Heřman et al (2017)
83.3%, Sp was 97%, positive predictive value- PPV was showed in their study that the coefficient of stiffness
83.3%, negative predictive value - NPV was 97%, and variability (CSV) which represents the ratio between the
the diagnostic accuracy was 94% (fig 6) [11]. maximum and minimum stiffness values recorded with
Klintworth et al (2012) showed that parotid tumors SWE is the strongest predictor of malignancy and allows
may have certain elastographic patterns: garland sign (re- parotid tumor differentiation between benignity and ma-
ticular distribution of stiff tissue in the whole tumor) for lignancy (p <0.0001) [27].
malignant tumors, dense core sign (very stiff central zone Regarding the modern US techniques, the use of elas-
with softer neighboring tissue) - characteristic of PMAs tography in the differential diagnosis of parotid tumors
and half-half sign (a stiff area located in the superficial is still controversial due to the overlap between the pa-
half of the lesion while the deeper part has a softer as- rameters. An important overlap between the values of
pect) – for WT [40]. elasticity of PMA and malignant tumors was reported in
In a recent study, Cantisani et al showed that benign the literature [10,38]. Moreover, adenolymphomas show
tumors had a mean elasticity contrast index (ECI) of 2.31 a great variety of histological structure, which is reflect-
and malignant tumors had a mean value of 5.55. The ECI ed in the wide range of SWE elasticity modulus of these
value >3.5 was the cutoff point for the differentiation of tumors reported by Huang et al (2018) [44]. However,
malignancies (90.5% diagnostic accuracy). The authors most authors accept that, in general, malignant tumors
concluded that USE with ECI index improves US Se in are stiffer than benign lesions, and PMAs are stiffer than
discriminating parotid tumors [20]. WTs, but this is not a strong enough criterion to allow the
Virtual touch imaging quantification (VTIQ) rep- differentiation of parotid tumors [10,11,26,28]. Zhang
resents a two-dimensional new form of shear wave im- et al (2019) in a meta-analysis presented a Sp of 67%
aging which allows combining quantitative and relative and a Se of 64% recorded by USE for the differentiation
stiffness imaging in a single screen. The investigation of parotid tumors [34]. Many authors have investigated
also allows the user to draw a two-dimensional region of methods to improve USE diagnostic accuracy in differ-
interest (ROI). VTIQ includes several share wave modes entiating parotid tumors. Thus, Mansour et al. (2019)
such as: velocity, quality, travel time, and displacement. [29] succeeded in differentiating PMA from WT using
VTIQ in combination with routine US techniques pro- conventional US together with SWE. Cantisani et al
vides additional data useful in distinguishing benign (2017) [20] using a cutoff value >3.5 for the elasticity
from malignant parotid tumors [41,42]. contrast index obtained significant results in differentiat-
Zengel et al (2017) using VTIQ showed that malig- ing malignant parotid tumors from benign lesions, with
nancies had much lower quality and the share wave ve- an accuracy of 90.5%.
locity was higher than 6.8 m/s in comparison with be- Herman et al (2017) obtained encouraging results
nign lesions [41]. Using the same investigation, Liu et a. for differentiating parotid tumors with the help of a co-
(2018) obtained higher values of share wave velocity for efficient of stiffness variability (CSV) [27]. Cortcu et al
malignant tumors compared to benign ones; the cutoff (2018) managed to differentiate between malignancy and
value was 2,445 m/s with a Sp of 80% and a Se of 91.8% benignity using USE based on strain ratio cutoff value of
[31]. Altinbas et al (2017) looking for the diagnostic 2.1 [11]. Encouraging results in the evaluation of parotid
value of SE in parotid tumors differentiation presented tumors with the help of USE have been obtained also by
the following results: mean elasticity score for benign other authors [11,31,39], which shows that this investiga-
Med Ultrason 2021; 23(2): 194-202 199
tion has a potential in differentiating parotid tumors that a significantly stronger perfusion enhancement with in-
must be investigated by further studies. USE, especially creased Doppler signal area. The authors concluded that
SE is an operator-dependent investigation, which may the Doppler signal area can show different courses de-
explain the large variations between the results of the pending on tumor histology [46].
published studies. Fischer et a. (2010) were successful in differentiating
PAM from WT based on the TTP parameter (26.8±11.1
Contrast-enhanced ultrasound versus 22.6±5.1 sec, p<0.05). Also, following the AUC
analysis, the PMA registered much lower values com-
CEUS allows a quantitative analysis of the microvas- pared to the WT [16].
cular perfusion of tumoral tissue, representing a reproduc- According to Klotz et al (2013) malignant tu-
ible measurement of perfusion kinetics after intravenous mors register prolonged MTT (17.94±1.62 s versus
administration of the contrast agent. It is a non-invasive 14.86±0.65 s; p<0.05) and AUC values (584.9±143 ver-
investigation with all the advantages of conventional US. sus 400.62±53.85; p<0.05) compared to benign lesions.
The contrast agent has a good safety profile, with a low Regarding the benign tumors, WT showed higher AUC
risk of side effects, and can be administered even to pa- values than PMA (515.4± 1.26 versus 285.82±36.44;
tients with impaired renal function due to its elimination p<0.05) [25].
through the respiratory tract (sulphur hexafluoride gas). Klotz et al (2014) using a new quantification software
The standard CEUS protocol involves injecting 4.8 ml of (Vuebox) for CEUS, were able to differentiate malignant
contrast agent into a peripheral vein followed by dynam- tumors from benign ones (p<0.05) as well as WT from
ic monitoring of the contrast agent passage through the PMA (p<0.05) based on the following parameters: AUC
ROI for minimum 120 seconds. The conventional gray- - 528.6±183.3 vs. 174.4±52.9 – malign vs. benign and
scale US images are displayed on the screen simultane- 302.8±36.2 – WT vs 65.6±13.8 – PMA, peak enhance-
ously with CEUS images. Quantitative parameters such ment - 155.3±69.5 vs. 38.2±9.0 – malign vs. benign and
as time-intensity curves (TIC) and enhancement patterns, 68.8±11.4 – WT vs 12.2±1.8 – PMA, wash-in-rate and
but also a number of semi-quantitative parameters such wash-in perfusion index. On the other hand, they did not
as area under the curve (AUC), time to peak (TTP) and find significant differences for mean transit time (MTT)
mean transit time (MTT) can be evaluated using CEUS and rise time (RT) [47].
investigation [8,10,25]. Küstermeyer et al (2016) using an innovative method
Wei et al (2013) assessed the morphologic and dis- of analysis the perfusion characteristics in 8 ROIs based
tribution features of micro-vascularity of the tumors and on CEUS, obtained a significant difference between be-
classified CEUS imaging into three types: type 1 - dif- nign parotid tumors compared to malignant ones [48]. On
fuse homogeneous enhancement; type 2 - heterogeneous the other hand, Badea et al (2013) could not find signifi-
enhancement with 3 subdivisions: a - more than 50% cant differences between the circulatory bed of benign
enhanced areas with well-defined margin, b - dotted or and malignant tumors using CEUS investigation [12].
splattered vessels or less than 50% enhanced areas with The CEUS parameters - TIC, TTP, MTT and AUC
well-defined margin and c - heterogeneous enhanced, le- were successful in the differential diagnosis of parotid
sions with ill-defined margin; and type 3 - no enhance- tumors [5,16,25]. Klotz et al (2013) showed in their study
ment and iso-enhancement with 2 subtypes: a - no en- that MTT and AUC showed significantly higher values
hancement in lesions and b - iso-enhancement between for malignant tumors compared to benign ones. The
lesions and surrounding tissues. The authors concluded circulation of the contrast agent through abnormal and
that types 1 and 3 are representative for benign tumors and disordered vessel architecture of cancer lesions may ex-
type 2, especially subtype c can suggest malignancy [45]. plain these increased values; on the other hand, it seems
Many studies published in the literature have shown that the reduced perfusion of benign tumors may occur
improved results with the using of CEUS for the differen- due to the regular architecture of the vascular systems of
tial diagnosis of parotid tumors (fig 7) [5,16,25,32]. these tumors [25]. Klotz et al (2014) [47] with the help
Data from the literature show that the use of multi- of a new software and the parameters such as: AUC, PE,
modal US to differentiate parotid tumors increased the WiR, WiPI manage to differentiate malignant parotid tu-
Se of malignant tumors detection from 77% to 91% but mors from benign ones with CEUS (fig 8).
decreased the Sp from 98% to 81% compared to conven- The preoperative differential diagnosis of parotid tu-
tional US [32]. mors based on US investigations is challenging for clini-
Steinhart et al (2003) showed that using US contrast cians due to the great histological variety of parotid tu-
agents in combination with Doppler US, PMA recorded mors. According to the latest World Health Organization
200 Sebastian Stoia et al Ultrasonography techniques in the preoperative diagnosis of parotid gland tumors
classification from 2017 [1], 20 malignant and 11 benign diagnosis of parotid tumors contribute to the controver-
tumor histology are described, to which another 9 histo- sies of this topic. In cases of inconclusive results, addi-
logical tumor types are added. The histological and US tional investigations and/or a second imaging opinion
characteristics that overlap between many tumor entities, could help establish the correct diagnosis.
as well as the lack of clear guidelines for the differential The use of MPUS provides promising results for pa-
rotid tumors differential diagnosis. Data from the litera-
ture shows high Se of up to 91% and Sp by combining
USE, CEUS and conventional US in differentiating pa-
rotid tumors; however, US still cannot replace the MRI
scan [32,49]. Further studies are necessary in order to
define the diagnostic value of the MPUS and to draw
clear indications of the method in the current diagnosis
of parotid tumors.
Based on the literature review [4-5,7-10,13,16-
20,25,28,30-32,45], a guide for the US differential di-
agnosis of parotid tumors is proposed by our group and
presented in fig 9.
Department of Ultrasound, Affiliated Dongyang Hospital of Wenzhou Medical University, Dongyang, Zhejiang,
China
Abstract
Aims: Foot and ankle injuries are a common presenting complaint in the emergency department. The diagnosis of foot
and ankle fractures is conventionally accomplished through X-rays. Whether ultrasound (US) can be considered as a primary
scanning modality is still a controversial issue; therefore, we did a meta-analysis to synthesize the diagnostic performance of
ultrasound for foot and ankle fractures. Material and methods: A comprehensive search was carried out to identify studies
in which patients with clinically suspected foot and ankle fractures were assessed by US. Two investigators independently
screened the literature and extracted the data. Any discrepancies were resolved via discussion. Study quality was assessed by
the Quality Assessment of Diagnostic Accuracy Studies 2 tool, and pooled sensitivity and specificity of various US findings
were determined. Results: Ten studies with a total of 1065 patients were included. There was significant heterogeneity across
the included studies. The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds
ratio for the diagnosis of foot and ankle fractures by US were 0.96 (95% confidence interval [CI], 0.90-0.99), 0.94 (95% CI,
0.88-0.97), 15.0 (95% CI, 7.9-28.6), 0.04 (95% CI, 0.02-0.11), and 367 (95% CI, 101-1338), respectively. Furthermore, the
summary receiver operating characteristic area under the curve was calculated to be 0.99. Conclusions: Ultrasound has an
excellent diagnostic performance for foot and ankle fractures and should be considered as a primary and radiation-free scan-
ning modality in the diagnosis of foot and ankle fractures.
Keywords: ultrasound; foot and ankle fractures; meta-analysis; systematic review; diagnostic accuracy
pediatric elbow, distal radius, metacarpal, phalanx, ankle foot and ankle fractures; (3) the accuracy of ultrasonog-
and metatarsal fractures [9–12]. One of the main advan- raphy in the diagnosis of foot and ankle fractures was
tages of ultrasound is the lack of ionizing radiation. Fur- evaluated and (4) a reference standard was adopted to
thermore, ultrasound can be immediately performed and confirm foot and ankle fractures, including X-ray, com-
is easily accessible in the ED, reducing diagnostic delays puted tomography, and/or magnetic resonance imaging.
and the time to the initiation of management. Additional The exclusion criteria were as follows: (1) case re-
advantages of ultrasound include the relative easiness to ports, letters, guidelines, consensus statements, and un-
teach [13], reduced pain experience [14], repeatability, published articles; (2) studies that contained an over-
portability, and the provision of additional information lapped population and (3) studies without sufficient data
regarding the musculoskeletal system [15–17]. to construct diagnostic 2x2 tables.
Whether ultrasound can be considered as a primary Data extraction
scanning modality is still a controversial issue, as the Two researchers (JW and YW) independently ex-
diagnostic accuracy of ultrasound for detecting foot and tracted the relevant data from the included studies us-
ankle fractures is variable across different studies. Prior ing a pre-designed data collection form. Any discrepan-
studies assessing the accuracy of US in the diagnosis of cies were resolved via discussion with the senior author
foot and ankle fractures have been published, with the (ZW). For eligible studies, the following items were ex-
sensitivity ranging from 83% to 100% and the specificity tracted: last name of the first author, year of publication,
ranging from 76% to 100% [18–20]. To our knowledge, country, study type, study setting, blinding method, US
no studies have comprehensively evaluated the literature equipment, probe frequency, sample size, number with
on foot and ankle fractures diagnosis using ultrasound. fractures, fracture prevalence, fracture site, mean age,
Hence, we did a meta-analysis to synthesize the diagnos- gender, US operator specialty, examiner training, US
tic performance of ultrasound for foot and ankle fractures. diagnostic criteria, standard reference, time between ul-
trasonography and the standard reference, true positives,
Material and methods true negatives, as well as false positives and false nega-
tives of US in the diagnosis of foot and ankle fractures.
Meta-analysis principles Study quality assessment
This meta-analysis was conducted according to the The Quality Assessment of Diagnostic Accuracy
Preferred Reporting Items for Systematic Reviews and Studies-2 (QUADAS-2) tool [22] was utilized to evaluate
Meta-Analyses (PRISMA) guidelines, which include 27 the risk of bias and methodological quality. The quality
items and provide specific guidance for the reporting of of each included study was evaluated by an appraisal of
systematic reviews [21]. the risk of bias of four domains and clinical applicability
Search strategy of three domains of the study characteristics. Four do-
Pubmed, EMBASE and Cochrane Library were sys- mains consisted of patient selection, index test, reference
tematically searched to identify potentially eligible stud- standard and flow and timing. Each domain was evalu-
ies from inception to March 2020. Computer searches ated for risk of bias, and the first three domains were
were carried out using the Medical Subject Heading and evaluated for applicability. The processing of the quality
keywords. Detailed search terms are provided in supple- assessment was performed utilizing RevMan 5.3 soft-
mentary file 1. The bibliographies of identified studies ware (Nordic Cochrane Centre, Copenhagen, Denmark).
and review articles were manually screened to expand Statistical analysis
the number of eligible studies. Only studies in English, The present meta-analysis was conducted by Stata
which satisfied the inclusion criteria, were included. 12.0 (Stata Corporation, College Station, Texas). All sta-
Inclusion and exclusion criteria tistical analyses were performed by one investigator, who
Two researchers (JW and YW) independently has experience in performing meta-analysis. The sum-
screened the titles and the abstracts of the potentially eli- mary estimates of sensitivity, specificity, positive likeli-
gible studies. Before identifying the literature, inclusion hood ratio (PLR), negative likelihood ratio (NLR) and
and exclusion criteria were defined to increase validity diagnostic odds ratio (DOR) with corresponding 95%
and reproducibility. Any disagreements between the two confidence intervals (CIs) were calculated using a bivari-
researchers were resolved via discussion with the senior ate random effect model in the present analysis, which
author (ZW). indicated the accuracy of US in the diagnosis of foot and
The inclusion criteria were as follows: (1) randomised ankle fractures. Meanwhile, the summary receiver op-
control trials and prospective studies were included; erator curve (SROC) was constructed and the area under
(2) studies involving patients with clinically suspected the curve (AUC) was calculated. An AUC close to 0.5
Med Ultrason 2021; 23(2): 203-212 205
shows a poor test, while an AUC of 1.0 demonstrates an viewed 1217 titles and abstracts and then excluded 1176
excellent diagnostic test [23]. We applied the spearman studies because it was obvious from the title or abstract
correlation analysis to determine whether a threshold ef- that they were not relevant to this meta-analysis. Full text
fect is present, with p < 0.05 representing a threshold ef- of the remaining studies was reviewed, and another 41
fect. The Cochrane Q test and the inconsistency index studies were excluded. Finally, 10 original research stud-
(I2) were used to assess the heterogeneity among differ- ies were included in the present meta-analysis [18-20,26-
ent studies with a p-value < 0.1 or I2 > 50% considered 32]. Manual searching of the reference cited in these 10
significant for heterogeneity [24]. studies did not yield any additional relevant studies. Fig-
Meta-regression analyses utilizing several covari- ure 1 shows a flow diagram summarizing the literature.
ates were carried out to investigate the potential causes Characteristics of the included studies
of heterogeneity: country (Turkey versus countries other The 10 included studies with a total of 1065 patients
than Turkey), sample size (>100 versus ≤100), fracture were published between 2009 and 2019 and written in
prevalence (>30% versus ≤30%), year published (2009- English. Five studies were conducted in Turkey [20,28-
2013 versus 2014-2019), reference standard (only X-ray 31], 1 was performed in France [18], 1 in Israel [19], 1
versus including CT or MRI), ultrasonographic operator in England [26], 1 in Sweden [27], and 1 in Iran [32]. All
(emergency physician versus others), ultrasonographic were prospective observational studies. The number of
training (yes versus others), blinding method (double patients in the study ranged from 37 to 246; 48.2% of pa-
blinding versus others), and bedside ultrasound (yes ver- tients were male and the mean age ranged from 8.1 to 52.7
sus others). The Deeks’ funnel plot asymmetry test was years. The prevalence of foot and ankle fractures ranged
applied to assess publication bias [25], through a p value from 10%-40.4%. Nine studies [19,20,26-32] were con-
>0.05 denoting no significant publication bias. ducted in ED and 1 was performed in the rheumatology
department [18]. Emergency physicians performed the US
Results examination in 5 studies [20,28,30-32], an experienced
rheumatologist in 1 study [18], a pediatric radiologist in
Study selection one study [19], an orthopedic surgeon in one study [27], an
The initial search for studies which assessed the di- ED member in 1 study [26] and a sonographer in 1 study
agnostic performance of US for foot and ankle fractures [29]. Double blinding between the standard reference and
provided 1818 studies, of which 1217 relevant studies index tests was found in 7 studies [18,20,26,27,29,31,32],
remained after removing 601 duplicate studies. We re- single blinding of the standard reference to US results in
2 studies [19,28] and 1 study [30] did not report a blind-
ing method. In all studies the time interval between US
and the standard reference was not given except for the
study by Banal et al who declared US and the standard
reference were performed on the same day [18]. Three
studies included ankle fractures [19,27,31], 3 included
foot or ankle fractures [26,28,29] and 4 only included
metatarsal fractures [18,20,30,32]. Seven studies used
the X-ray as the reference standard [19,20,26-28,30,32],
1 used MRI [18] and 2 used X-ray or CT [29,31]. The
probe frequency ranged from 5 to 15 MHz. Table I and
II epitomizes the data extracted from the included stud-
ies. More details are showed in supplementary file 2.
Quality assessment
with values ranging from 85.9 to 100% and 86.4 to 100% fractures [19,20,26-32]. Other ultrasonic findings such
respectively [11]. This is consistent with the results of as hypoechoic periosteal elevation, swollen soft tissues
our study. and increased vascularity, which are often not visualized
As we all know, ultrasound is an operator-dependent by radiography also support foot and ankle fractures in-
technology [36]. Therefore, it is of the highest impor- directly [18]. So further prospective studies with larger
tance to ensure that operators acquire sufficient train- sample sizes are required to identify a well-defined pro-
ing and practice with this technology. With regard to the tocol for ultrasound in detecting foot and ankle fractures.
subgroup analysis of US training, operators with extra Significant heterogeneity had been observed in this
ultrasonographic training had a comparable diagnostic meta-analysis (sensitivity: I2 = 51.43%, p=0.03; specific-
performance to others such as experience operators or ity: I2 = 81.45%, p=0.00) and meta-regression analyses
not reported (sensitivity: 0.96 and 0.96; specificity: 0.94 showed that the blinding method accounted for part of
and 0.92). However, the training protocols were vari- the significant sources of heterogeneity in terms of sen-
able among studies and the most of the training courses sitivity. However, there were other factors which might
ranged from 30 minutes to 2 days [26,27,29-31]. It is un- involve the significant heterogeneity. Other factors such
defined what the optimal training protocol is, so further as specialties of ultrasonic operators, different experi-
prospective studies are required to determine the optimal ence levels and different equipment might also play an
training protocol and learning curve for this technology. important role in heterogeneity among studies. However
According to the eligible studies, the main ultrasonic meta-regression analyses could not be performed to ex-
characteristics of foot and ankle fractures included cor- plore the significant sources of heterogeneity according
tical disruption or stepping, axial deviation of the bone to other factors referred above because of the insufficient
surface and cortical depression, which directly indicate information in the included studies.
210 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
Most of the included studies chose conventional the X-ray, the quantity of radiographs needed would have
X-ray as the reference standard, which is not the golden fallen by 80.9% [26]. Consequently, US is increasingly
standard to detect fractures. Because occult fractures ac- being considered as a first-line modality in the primary
count for 2%-36% in a conventional X-ray on account response to emergency situations and could decrease the
of overlapping structures, under-mineralized ossification need for radiographic imaging in patients with foot and
centres and non-perpendicular X-ray beam to the fracture ankle fractures.
line [35]. Several studies have demonstrated that multi- It is important to consider some limitations with re-
planar capabilities of ultrasound might make it superior spect to this study. First, a relatively small number of
to radiography in the detection of occult fractures that the studies were included in the present meta-analysis as a
X-ray has missed [37,38]. Banal et al found that in cases result of the limited relevant high-quality studies, and the
of normal radiographs, US is indicated in the diagnosis literature search merely included studies written in Eng-
of metatarsal bone stress fractures with a sensitivity of lish. Anyway, we were able to acquire several important
0.83 and a specificity of 0.76 [18]. The prospective study conclusions with respect to the diagnostic value of US
of ultrasonographic evaluation of radiographically nega- and related factors. Second, most of the eligible studies
tive ankle injuries in a pediatric population performed failed to report the precise duration between the stand-
by Simanovsky et al demonstrated that ultrasound is ef- ard reference and US examination except for the study
fective in the detecting radiographically silent fractures by Banal et al who declared that US and dedicated MRI
of the pediatric ankle with a sensitivity of 1.00 and a examinations of the metatarsal bones were performed the
specificity of 0.97 [19]. In future, studies adopting CT or same day [18]. However, the reference standard and US
MRI as the golden standard are required to evaluate the performed without a narrow time frame may not increase
diagnostic performance of ultrasound in detecting radio- the performance bias because fractures will not change
graphically occult fractures. over time in the short term. Third, no study evaluated
Atilla et al showed that the sensitivity and specific- intraobserver or interobserver variability; however, it is
ity of US for foot and ankle fractures could vary accord- important to do so because ultrasound is an operator-
ing to the fracture site. US had excellent sensitivity and dependent modality. Therefore, more further studies are
specificity in the diagnosis of the fifth metatarsal frac- required to evaluate the intraobserver or interobserver
tures with a sensitivity of 1.00 and specificity of 0.96, variability. Finally, most of the included studies had
which might be explained by the superficial site of the methodological limitations, especially in domains such
fifth metatarsal and its smooth contours and ease of view- as patient selection, the index test, reference standard and
ing from different planes by US; while the sensitivity and flow and timing, and therefore improvements in the fu-
specificity of US for detecting navicular fractures was ture study design are required to accurately address the
relatively lower (sensitivity: 0.40 and specificity: 0.93) issue under investigation.
as the dorsal surface of the navicular bone is irregular and
could only be viewed in the dorsal plane [29]. Our sub- Conclusions
group analysis of the 4 studies only including metatarsal In summary, this comprehensive meta-analysis dem-
fractures [18,20,30,32] showed pooled sensitivity of 0.95 onstrates that ultrasound has an excellent diagnostic
and specificity of 0.90. However, we faced insufficient performance for foot and ankle fractures and should be
data to perform more subgroup analyses regarding dif- considered as a primary and radiation-free scanning mo-
ferent fracture sites such as ankle fractures or navicular dality in the diagnosis of foot and ankle fractures. How-
fractures. ever, the conclusion of this meta-analysis on the strength
Foot and ankle injuries are almost universally as- of a small quantity of studies that met the specific in-
sessed by the Ottawa Foot and Ankle Rules, which have clusion criteria should be interpreted with caution. Large
a reported sensitivity of 97.9% to 99.8% and a specificity prospective international multicenter studies are still re-
of 28.8% and 42.3% for foot and ankle fractures respec- quired to support the present conclusion.
tively [39-42]. The poor specificity of the rules indicates
that about 60% of patients who undergo a radiograph Conflict of interest: none
do not have a fracture, with many patients exposed to
unnecessary harmful ionizing radiation. Hedelin et al References
suggested ultrasound-guided triage seemed to be able to 1. Shibuya N, Davis ML, Jupiter DC. Epidemiology of foot
decrease the need for radiographic imaging in patients and ankle fractures in the United States: an analysis of the
with ankle trauma [27]. Similarly, Canagasabey et al sug- National Trauma Data Bank (2007 to 2011). J Foot Ankle
gested that US examination had been employed prior to Surg 2014;53:606–608.
Med Ultrason 2021; 23(2): 203-212 211
2. Crowley SG, Trofa DP, Vosseller JT, et al. Epidemiology of sectional study from Shanxi Province, China. Medicine
foot and ankle injuries in National Collegiate Athletic As- (Baltimore) 2018;97:e13230.
sociation men’s and women’s ice hockey. Orthop J Sports 18. Banal F, Gandjbakhch F, Foltz V, et al. Sensitivity and spec-
Med 2019;7:2325967119865908. ificity of ultrasonography in early diagnosis of metatarsal
3. Wood JN, Henry MK, Berger RP, et al. Use and utility of bone stress fractures: a pilot study of 37 patients. J Rheu-
skeletal surveys to evaluate for occult fractures in young matol 2009;36:1715–1719.
injured children. Acad Pediatr 2019;19:428–437. 19. Simanovsky N, Lamdan R, Hiller N, Simanovsky N. So-
4. Nery C, Raduan F, Baumfeld D. Foot and ankle injuries in nographic detection of radiographically occult fractures
professional soccer players: diagnosis, treatment, and ex- in pediatric ankle and wrist injuries. J Pediatr Orthop
pectations. Foot Ankle Clin. 2016;21:391–403. 2009;29:142–145.
5. Ait-Ali L, Andreassi MG, Foffa I, Spadoni I, Vano E, Pi- 20. Yesilaras M, Aksay E, Atilla OD, Sever M, Kalenderer O.
cano E. Cumulative patient effective dose and acute radi- The accuracy of bedside ultrasonography as a diagnostic
ation-induced chromosomal DNA damage in children with tool for the fifth metatarsal fractures. Am J Emerg Med
congenital heart disease. Heart 2010;96:269–274. 2014;32:171–174.
6. Brenner DJ, Doll R, Goodhead DT, et al. Cancer risks at- 21. Shamseer L, Moher D, Clarke M, et al. Preferred report-
tributable to low doses of ionizing radiation: assessing what ing items for systematic review and meta-analysis proto-
we really know. Proc Natl Acad Sci U S A 2003;100:13761– cols (PRISMA-P) 2015: elaboration and explanation. BMJ
13766. 2015;350:g7647.
7. LiMarzi GM, Scherer KF, Richardson ML, et al. CT and 22. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2:
MR imaging of the postoperative ankle and foot. Radio- a revised tool for the quality assessment of diagnostic ac-
graphics 2016;36:1828–1848. curacy studies. Ann Intern Med 2011;155:529–536.
8. Beard NM, Gousse RP. Current ultrasound application in 23. Hanley JA, McNeil BJ. The meaning and use of the area
the foot and ankle. Orthop Clin North Am 2018;49:109– under a receiver operating characteristic (ROC) curve. Ra-
121. diology 1982;143:29–36.
9. Lee SH, Yun SJ. Diagnostic performance of ultrasonogra- 24. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Meas-
phy for detection of pediatric elbow fracture: a meta-analy- uring inconsistency in meta-analyses. BMJ 2003;327:557–
sis. Ann Emerg Med 2019;74:493–502. 560.
10. Douma-den Hamer D, Blanker MH, Edens MA, et al. Ultra- 25. Deeks JJ, Macaskill P, Irwig L. The performance of tests
sound for distal forearm fracture: a systematic review and of publication bias and other sample size effects in system-
diagnostic meta-analysis. PLoS One 2016;11:e0155659. atic reviews of diagnostic test accuracy was assessed. J Clin
11. Champagne N, Eadie L, Regan L, Wilson P. The effective- Epidemiol 2005;58:882–893.
ness of ultrasound in the detection of fractures in adults 26. Canagasabey MD, Callaghan MJ, Carley S. The sono-
with suspected upper or lower limb injury: a systematic graphic Ottawa Foot and Ankle Rules study (the SOFAR
review and subgroup meta-analysis. BMC Emerg Med study). Emerg Med J 2011;28:838–840.
2019;19:17. 27. Hedelin H, Goksör LÅ, Karlsson J, Stjernström S. Ultra-
12. Zhao W, Wang G, Chen B, et al. The value of ultrasound for sound-assisted triage of ankle trauma can decrease the need
detecting hand fractures: A meta-analysis. Medicine (Balti- for radiographic imaging. Am J Emerg Med 2013;31:1686–
more) 2019;98:e17823. 1689.
13. Mircea PA, Badea R, Fodor D, Buzoianu AD. Using ul- 28. Ekinci S, Polat O, Günalp M, Demirkan A, Koca A. The
trasonography as a teaching support tool in undergraduate accuracy of ultrasound evaluation in foot and ankle trau-
medical education - time to reach a decision. Med Ultrason ma. Am J Emerg Med 2013;31:1551–1555.
2012;14:211–216. 29. Atilla OD, Yesilaras M, Kilic TY, et al. The accuracy of
14. Saul T, Ng L, Lewiss RE. Point-of-care ultrasound in the bedside ultrasonography as a diagnostic tool for fractures
diagnosis of upper extremity fracture-dislocation. A picto- in the ankle and foot. Acad Emerg Med 2014;21:1058–
rial essay. Med Ultrason 2013;15:230–236. 1061.
15. Sconfienza LM, Albano D, Allen G, et al. Clinical indica- 30. Kozaci N, Ay MO, Avci M, et al. The comparison of ra-
tions for musculoskeletal ultrasound updated in 2017 by diography and point-of-care ultrasonography in the di-
European Society of Musculoskeletal Radiology (ESSR) agnosis and management of metatarsal fractures. Injury
consensus. Eur Radiol 2018;28:5338–5351. 2017;48:542–547.
16. Zayat AS, Md Yusof MY, Wakefield RJ, Conaghan PG, 31. Ozturk P, Aksay E, Oray NC, Bayram B, Basci O, Tokgoz
Emery P, Vital EM. The role of ultrasound in assessing D. Emergency physician accuracy using ultrasonography
musculoskeletal symptoms of systemic lupus erythemato- to diagnose lateral malleolar fracture. Am J Emerg Med
sus: a systematic literature review. Rheumatology (Oxford) 2018;36:362–365.
2016;55:485–494. 32. Ebrahimi M, Habibzadeh SR, Ahmadi SR, Khajeh Nasiri
17. Zhang CM, Zhang JF, Xu J, Guo YL, Wang G, Yang LH. S, Kaveh MM, Foroughian M. Diagnostic Accuracy of Ul-
Musculoskeletal ultrasonography for arthropathy assess- trasonography in Diagnosis of Metatarsal Bone Fracture; a
ment in patients with hemophilia: A single-center cross- Cross Sectional Study. Arch Acad Emerg Med 2019;7:e49.
212 Jiangfeng Wu, Yunlai Wang et al The diagnostic accuracy of ultrasound in the detection of foot and ankle fractures
33. Gordon I, Sinert R, Chao J. The utility of ultrasound in 38. Allen GM, Wilson DJ, Bullock SA, Watson M. Extrem-
detecting skull fractures after pediatric blunt head trauma: ity CT and ultrasound in the assessment of ankle inju-
systematic review and meta-analysis. Pediatr Emerg Care ries: occult fractures and ligament injuries. Br J Radiol
2020, doi:10.1097/PEC.0000000000001958. 2020;93:20180989.
34. Chartier LB, Bosco L, Lapointe-Shaw L, Chenkin J. Use of 39. Bachmann LM, Kolb E, Koller MT, Steurer J, ter Riet G.
point-of-care ultrasound in long bone fractures: a system- Accuracy of Ottawa ankle rules to exclude fractures of the
atic review and meta-analysis. CJEM 2017;19:131–142. ankle and mid-foot: systematic review. BMJ 2003;326:417.
35. Douma-den Hamer D, Blanker MH, Edens MA, et al. Ultra- 40. Auleley GR, Ravaud P, Giraudeau B, et al. Implementation
sound for distal forearm fracture: a systematic review and of the Ottawa ankle rules in France. A multicenter rand-
diagnostic meta-analysis. PLoS One 2016;11:e0155659. omized controlled trial. JAMA 1997;277:1935–1939.
36. Scheel AK, Schmidt WA, Hermann KG, et al. Interobserver 41. Lucchesi GM, Jackson RE, Peacock WF, Cerasani C,
reliability of rheumatologists performing musculoskeletal Swor RA. Sensitivity of the Ottawa rules. Ann Emerg Med
ultrasonography: results from a EULAR “Train the train- 1995;26:1–5.
ers” course. Ann Rheum Dis 2005;64:1043–1049. 42. Beckenkamp PR, Lin CC, Macaskill P, Michaleff ZA, Ma-
37. Kwee RM, Kwee TC. Ultrasound for diagnosing ra- her CG, Moseley AM. Diagnostic accuracy of the Ottawa
diographically occult scaphoid fracture. Skeletal Radiol Ankle and Midfoot Rules: a systematic review with meta-
2018;47:1205–1212. analysis. Br J Sports Med 2017;51:504–510.
Review Med Ultrason 2021, Vol. 23, no. 2, 213-219
DOI: 10.11152/mu-2727
1HeartInstitute “N. Stancioiu”, 2Pulmonology Department, 32nd Internal Medicine Department, “Iuliu Hatieganu”
University of Medicine and Pharmacy Cluj Napoca, Romania
Abstract
Despite efforts to treat obstructive sleep apnea syndrome (OSA), the condition remains an important risk factor in the
development of cardiovascular disease. Early detection of RV dysfunction with novel echocardiographic techniques (speckle
tracking echocardiography) may be useful in preventing progression to pulmonary hypertension, with subsequent heart failure
and cardiovascular death. Echocardiography is the method of choice for the evaluation of OSA consequences on the heart.
Although standard echocardiographic parameters are routinely used in these patients, there are several limitations in the early
detection of RV dysfunction. The main concerns are the complex geometry of RV and the impact of pre- and afterload on RV
myocardium, which cannot be assessed through standard measurements. The aim of this review is to highlight the utility of
advanced echocardiographic parameters in the identification of OSA patients with subclinical myocardial dysfunction, which
are at risk of developing heart failure and later adverse events. Speckle tracking echocardiography might provide higher sensi-
tivity in unmasking alterations in RV function when conventional echocardiographic methods cannot detect them. Therefore,
this method has a major role in the detection of early stages of RV dysfunction, along with better risk stratification and better
timing in the initiation of therapy.
Keywords: obstructive sleep apnea; speckle tracking; echocardiography
of 94.4 % and 64.7% respectively. The prognostic value who did not find any differences in stain or strain rate
of RV LS was better than other indicators of RV function at patients with OSA and pulmonary hypertension. Con-
used alone or in combination, and it was maintained in cerning RV strain in OSA patients during CPAP therapy,
univariate and multivariate Cox regression analysis [36]. free wall RV strains and global strain improved after
2DE speckle tracking of the RV CPAP treatment, even in the presence of unchanged con-
in patients with OSA ventional parameters [39]. Kim et al showed that CPAP
RV function plays an important role in the morbidity therapy, unlike sham therapy significantly improved RV
and mortality of patients with OSA. Early detection of GLS and also reduced RV dimension [43]. Other data re-
RV dysfunction before the onset of pulmonary hyperten- port strain improvement under CPAP therapy in apical
sion is crucial in preventing progression to heart failure segments, but not in the medial or basal ones, explain-
and cardiac death [37]. This became possible using 2D ing why RV global longitudinal strain (average of apical,
speckle tracking echocardiography. Studies showed that medial and basal segments) remained unchanged after 6
RV longitudinal strain had a predicted value in several months of CPAP therapy [42].
cardiovascular diseases (CHD, heart failure, valvular 3D speckle tracking in the evaluation of the RV:
heart disease, PH and OSA). RV strain in OSA patients a new era
has not been yet rigorously and exhaustively explored, The RV has a complex pattern of contraction, which
but there are investigations on this subject with encourag- occurs along three anatomically relevant axes: longitudi-
ing results. [38]. Recent studies have shown that patients nal shortening with tricuspid annulus traction towards the
with OSA have shown impairment of global longitudinal apex, anteroposterior shortening by stretching the free
strain (GLS-RV) and lateral wall strain (LS-RV) in the wall over the interventricular septum and radial move-
presence of preserved RV-EF. As Buonauro et al [39], ment of the free wall, also known as the “bellows effect“.
concluded in their study, OSA patients showed signifi- Although it is well known that the subepicardial layer of
cant reduction in RV GLS (18.2±2.4%, p<0.001) and LS- the myocardium is rich in circumferential myofibers, in
RV (17.8±4.8%, p<0.001), but not in septal longitudinal the daily echocardiographic evaluation of the right ven-
strain (SLS-RV) (18.8±3.6 %, p=0.229) compared with tricle, non-longitudinal motion directions are often omit-
healthy controls. RV-GLS correlated with OSA severity, ted [44]. Therefore, the mix of deformation marks pro-
in the absence of important alteration of other echocar- vides insight into the pathophysiologic mechanics of RV
diographic parameters of RV systolic function, such as dysfunction. In comparison to 2D STE, the advantage of
3DE-EF or TAPSE. Li et al [40] showed significantly using 3D STE consists of the possibility of simultaneous
reduced RV LS and strain rates of RV apical segments assessment of various wall motion indices, in particular:
in patients with mild OSA compared to controls, results radial strain (RS), longitudinal strain (LS) and circumfer-
confirmed by the study of Kepez et al [11]. D’Andrea ential strain (CS) in the entire RV myocardium. Another
et al [41], revealed a reduction in both LV and RV GLS unique parameter assessed by 3DE STE is the area change
(-13.8±5.2 %, p<0.01), as well as RV LS in OSA patients ratio (ACR) or area strain in a regional or entire track-
compared with controls. In concordance with these find- ing area [45]. The clinical utility and feasibility of the
ings, Hammerstingl et al [42], found that RV 2DE GLS 3D STE of RV technique were evaluated by Atsumi et al
(-14.5 ± 8.2 %, p<0.0001) was significantly lower in sub- [46], in both an experimental and clinical study, by using
jects with higher AHI, compared with controls; and was sonomicrometry crystals implanted in the RV endocar-
positively associated with the severity of OSA. More- dium, in order to validate 3D STE data. The authors ob-
over, conventional echocardiography parameters, such as tained significant correlation between the two techniques.
TAPSE, RV MPI, were not significantly impaired in these The PH group had significantly lower RV global defor-
patients, supposing that they are not sufficiently sensitive mation parameters than the control group. These findings
in detecting subclinical changes of RV function in OSA. confirm the results of previous studies that used 3D STE
RV dysfunction might be the expression of pulmonary modified for RV software, showing that PH patients have
hypertension. This hypothesis is supported by several reduced RV strain values, poorer outcomes, lowered
studies, which demonstrate that patients with OSA and ACR, LS and CS and increased mortality risk compared
right ventricular systolic pressure (RVSP) ≥30 mmHg, with controls [47]. On the other hand, Liu et al evalu-
show substantially lower GLS-RV (18.2±2.4%, p<0.001), ated and compared the value of RV LS detected by both
LLS-RV (17.8± 4.8%, p<0.001), compared with patients 2DE and 3D STE in the risk stratification of patients with
with RVSP<30 mmHg (GLS-RV=22.7±4.8%, LLS- pre-capillary PH. Authors demonstrated that 2D RV LS is
RV=25.1±7.0%) [28]. However, these data are in op- better than 3D RV LS for the detections of patients with
position with the results obtained by Altekin et al [37], pre-capillary hypertension intermediate-high risk [48].
Med Ultrason 2021; 23(2): 213-219 217
3D speckle tracking echocardiography for the
RV: are regional deformation parameters also
necessary?
The myofibers in the myocardium are a complex 3D
network, arranged as a multiple helical display with im-
portant functional effect. The orientation of these my-
ofibers is important because it helps to understand the
pattern of RV contraction in 3D space [42]. Knowing that
3D STE allows a better perception of RV wall dynamic/
motion, the question is which of the 3D strain parameters
is more dependable.
Atsumi et al [46] showed in their experimental study
(10 sheep), that RV ACR and RV LS were significantly Fig 2. Representative 3D RV strain images in two patients with
obstructive sleep apnea. To asses global and longitudinal RV
lower in moderate pulmonary artery banding (peak RV systolic function, we used a 6 segment RV model basal RV free
pressure>40 mmHg) and in severe pulmonary artery wall, mid RV free wall, apical RV wall, apical septum, mid sep-
banding (peak RV pressure >60 mmHg) compared to tum, and basal septum). In both subjects left ventricle function
baseline values; however, RV CS differed significantly was normal. a) 3D RV speckle-tracking bulls eye in a patient
with severe obstructive sleep apnea and PH (decrease of GPSL
only between baseline and sever PAB, revealing that
RV = -15). b) 3D RV speckle-tracking bulls eye in a patient with
longitudinal contraction may be more reactive to pres- mild to moderate obstructive sleep apnea without PH (GPSL
sure overload. The difference between LS and CS was RV = -28). GPSL, global peak longitudinal strain; PH, pulmo-
explained by the RV myofibers arrangement: the endo- nary hypertension; RV, right ventricular.
cardial layer contains longitudinal myofibers, the epicar-
dial layer comprises circumferential myofibers, whereas subclinical stages of RV dysfunction. Furthermore, it
the epicardial fiber is continuous with the left ventricu- might be useful in evaluating the efficacy of continuous
lar myocardium. Therefore, circumferential contraction positive air pressure therapy in improving cardiac func-
might be less affected by pressure overload solely in the tion. Thus, an earlier initiation of OSA therapy can re-
RV due to its preservation by LV contraction. duce cardiovascular risk.
On the other hand, in a study comparing 97 patients Although these data provide motivation for using the
suffering from PHT with 60 controls, Smith et al [47] more recent echocardiographic techniques for the assess-
demonstrated that all PHT patients had reduced RV ment of cardiac function in patients with OSA and their
strain indices compared with controls; RV AS and CS response to therapy, several issues need further investi-
were strongly correlated to RV EF (p<0.001). LS and RS gation in order to be accepted as mainstream methods
were also correlated to RV EF, but lesser. Moreover, the for the quantitative assessment of RV function. Speckle
authors evaluated mortality at 24 months, showing that tracking imaging is a more sensitive technique for detect-
RV AS, CS and LS predicted it most accurately. Further- ing subclinical alterations in ventricular functions that
more, only AS, and in a lesser level, age, were predictors contrarily may be missed by conventional echocardiog-
of death, underling the superiority of 3D STE RV – AS raphy. The favourite deformation parameter is the global
over other parameters. A possible explication for why longitudinal strain due to its reproducibility and feasibil-
RV CS may be more affected in the group of patients ity. 3DE is useful in the determination of ejection frac-
with PHT, is that important structural changes happen tion, RV volumes and intracardiac anatomy. It is strongly
during the development of PHT, including a change in recommended to develop and validate a dedicated echo-
fiber orientation and increased wall stress, leading to re- cardiography program for each specific OSA lesion and
duced circumferential contraction (fig 2). to combine deformation imaging with 3DE [49].
3D speckle tracking of RV in patients with OSA:
future directions Conclusion
The impact of OSA on the right ventricular structure
and function has been demonstrated, since PAH, impair- Deterioration of RV mechanics occurs before func-
ment of RV systolic function and RV dyssynchrony have tional and structural heart damage. Therefore, recent
an important influence on cardiovascular morbidity and echocardiographic techniques seem to be the cornerstone
mortality. 3DE speckle tracking allows a better and more of detecting subtle cardiac changes in the RV. Since RV
comprehensive assessment of the RV wall motion. More- demonstrated an important predictive value in several
over, this novel technique is able to detect alterations in pathologies, it should be more widely used in the evalu-
218 Ioana Maria Chetan et al Echocardiography in patients with obstructive sleep apnea and RV subclinical myocardial dysfunction
ation of patients with OSA, which might present early 13. Addetia K, Muraru D, Badano LP, Lang RM. New Direc-
changes in RV contraction, even in the presence of an tions in Right Ventricular Assessment Using 3-Dimensional
apparently preserved systolic function. Risk stratifica- Echocardiography. JAMA Cardiol 2019;4:936-944.
14. Narang A, Freed B. The Future of Imaging in Pulmonary
tion in these patients using modern echocardiography
Hypertension: Better Assessment of Structure, Function,
techniques might bring future information regarding the
and Flow. Adv Pulm Hypertens 2019;18:126-133.
impact on the cardiac function of the disease and might 15. Addetia K, Maffessanti F, Yamat M, et al. Three-dimen-
also offer guidance for early therapy in order to prevent sional echocardiography-based analysis of right ventricular
further damage on the systolic RV function. shape in pulmonary arterial hypertension. Eur Heart J Car-
diovasc Imaging 2016;17:564-575.
Conflict of interest: none 16. van der Zwaan H, Helbing W, McGhie J, et al. Clinical
Value of Real-Time Three-Dimensional Echocardiography
References for Right Ventricular Quantification in Congenital Heart
1. Dong R, Dong Z, Liu H, Shi F, Du J. Prevalence, Risk Fac- Disease: Validation With Cardiac Magnetic Resonance Im-
tors, Outcomes, and Treatment of Obstructive Sleep Apnea aging. J Am Soc Echocardiogr 2010;23:134-140.
in Patients with Cerebrovascular Disease: A Systematic Re- 17. Nagata Y, Wu VC, Kado Y, et al. Prognostic Value of
view. J Stroke Cerebrovasc Dis 2018;27:1471-1480. Right Ventricular Ejection Fraction Assessed by Tran-
2. Hersi AS. Obstructive sleep apnea and cardiac arrhythmias. sthoracic 3D Echocardiography. Circ Cardiovasc Imaging
Ann Thorac Med 2010;5:10. 2017;10:e005384.
3. Zhou NW, Pan CZ, Kong DH, et al. A novel method for 18. Tamborini G, Brusoni D, Torres Molina JE, et al. Feasi-
sensitive determination of subclinical right ventricular sys- bility of a New Generation Three-Dimensional Echocardi-
tolic dysfunction in patients with obstructive sleep apnea. ography for Right Ventricular Volumetric and Functional
Clin Respir J 2017;11:951-959. Measurements. Am J Cardiol 2008;102:499-505.
4. Lee CHK, Leow LC, Song PR, Li H, Ong TH. Acceptance 19. Renella P, Marx GR, Zhou J, Gauvreau K, Geva T. Feasibil-
and Adherence to Continuous Positive Airway Pressure ity and Reproducibility of Three-Dimensional Echocardio-
Therapy in patients with Obstructive Sleep Apnea (OSA) graphic Assessment of Right Ventricular Size and Function in
in a Southeast Asian privately funded healthcare system. Pediatric Patients. J Am Soc Echocardiogr 2014;27:903-910.
Sleep Sci 2017;10:57-63. 20. Lang RM, Badano LP, Mor-Avi V, et al. Recommendations
5. Bradley TD, Floras JS. Obstructive sleep apnoea and its for Cardiac Chamber Quantification by Echocardiography
cardiovascular consequences. Lancet 2009;373:82-93. in Adults: An Update from the American Society of Echo-
6. Broström A, Årestedt KF, Nilsen P, Strömberg A, Ulander cardiography and the European Association of Cardiovas-
M, Svanborg E. The side-effects to CPAP treatment inven- cular Imaging. J Am Soc Echocardiogr 2015;28:1-39.e14.
tory: the development and initial validation of a new tool 21. Vitarelli A, Terzano C, Saponara M, et al. Assessment of
for the measurement of side-effects to CPAP treatment. J Right Ventricular Function in Obstructive Sleep Apnea
Sleep Res 2010;19:603-611. Syndrome and Effects of Continuous Positive Airway Pres-
7. Shahar E, Whitney CW, Redline S, et al. Sleep-disordered sure Therapy: A Pilot Study. Can J Cardiol 2015;31:823-
breathing and cardiovascular disease: cross-sectional re- 831.
sults of the Sleep Heart Health Study. Am J Respir Crit Care 22. Shimada YJ, Shiota M, Siegel RJ, Shiota T. Accuracy of
Med 2001;163:19-25. Right Ventricular Volumes and Function Determined by
8. Stradling J, Davies RJ. Sleep apnea and hypertension— Three-Dimensional Echocardiography in Comparison with
what a Mess! Sleep 1997;20:789-793. Magnetic Resonance Imaging: A Meta-Analysis Study. J
9. Maripov A, Mamazhakypov A, Sartmyrzaeva M, et al. Am Soc Echocardiogr 2010;23:943-953.
Right Ventricular Remodeling and Dysfunction in Obstruc- 23. Jenkins C, Chan J, Bricknell K, Strudwick M, Marwick TH.
tive Sleep Apnea: A Systematic Review of the Literature Reproducibility of Right Ventricular Volumes and Ejection
and Meta-Analysis. Can Respir J 2017;2017: 1587865. Fraction Using Real-time Three-Dimensional Echocardiog-
10. Podszus T, Bauer W, Mayer J, Penzel T, Peter JH, Wichert raphy. Chest 2007;131:1844-1851.
P. Sleep apnea and pulmonary hypertension. Klin Wochen- 24. Knight DS, Grasso AE, Quail MA, et al. Accuracy and Re-
schr 1986;64:131-134. producibility of Right Ventricular Quantification in Patients
11. Kepez A, Niksarlioglu E, Hazirolan T, et al. Early Myo- with Pressure and Volume Overload Using Single-Beat
cardial Functional Alterations in Patients with Obstructive Three-Dimensional Echocardiography. J Am Soc Echocar-
Sleep Apnea Syndrome. Echocardiography 2009;26:388- diogr 2015;28:363-374.
396. 25. Kjaergaard J, Petersen CL, Kjaer A, Schaadt BK, Oh JK,
12. Kossaify A. Echocardiographic Assessment of the Right Hassager C. Evaluation of right ventricular volume and
Ventricle, from the Conventional Approach to Speckle function by 2D and 3D echocardiography compared to
Tracking and Three-Dimensional Imaging, and Insights MRI. Eur J Echocardiogr 2006;7:430-438.
into the “Right Way” to Explore the Forgotten Chamber. 26. Morcos P, Vick GW 3rd, Sahn DJ, Jerosch-Herold M, Shur-
Clin Med Insights Cardiol 2015;9:65-75. man A, Sheehan FH. Correlation of right ventricular ejec-
Med Ultrason 2021; 23(2): 213-219 219
tion fraction and tricuspid annular plane systolic excursion 38. Tadic M, Cuspidi C, Grassi G, Mancia G. Obstructive
in tetralogy of Fallot by magnetic resonance imaging. Int J sleep apnea and cardiac mechanics: how strain could help
Cardiovasc Imaging 2008;25:263-270. us? Heart Fail Rev 2020 Feb 3. doi:10.1007/s10741-020-
27. Zou H, Leng S, Xi C, et al. Three-dimensional biventricu- 09924-0
lar strains in pulmonary arterial hypertension patients using 39. Buonauro A, Galderisi M, Santoro C, et al. Obstructive
hyperelastic warping. Comput Methods Programs Biomed sleep apnoea and right ventricular function: A combined as-
2020;189:105345. sessment by speckle tracking and three-dimensional echo-
28. Mor-Avi V, Lang RM, Badano LP, et al. Current and cardiography. Int J Cardiol 2017;243:544-549.
evolving echocardiographic techniques for the quantita- 40. Li J, Wang Z, Li Y, et al. Assessment of regional right ven-
tive evaluation of cardiac mechanics: ASE/EAE consensus tricular systolic function in patients with obstructive sleep
statement on methodology and indications endorsed by the apnea syndrome using velocity vector imaging. Medicine
Japanese Society of Echocardiography. J Am Soc Echocar- 2016;95:e4788.
diogr 2011;24:277-313. 41. D’Andrea A, Martone F, Liccardo B, et al. Acute and
29. Kasikcioglu HA, Karasulu L, Tartan Z, Kasikcioglu E, Cu- Chronic Effects of Noninvasive Ventilation on Left and
hadaroglu C, Cam N. Occult cardiac dysfunction in patients Right Myocardial Function in Patients with Obstructive
with obstructive sleep apnea syndrome revealed by tissue Sleep Apnea Syndrome: A Speckle Tracking Echocardio-
Doppler imaging. Int J Cardiol 2007;118:203-205. graphic Study. Echocardiography 2016;33:1144-1155.
30. Dursunoğlu N, Dursunoğlu D, Kılıç M. Impact of obstruc- 42. Hammerstingl C, Schueler R, Wiesen M, et al. Impact of
tive sleep apnea on right ventricular global function: sleep Untreated Obstructive Sleep Apnea on Left and Right Ven-
apnea and myocardial performance index. Respiration tricular Myocardial Function and Effects of CPAP Therapy.
2005;72:278-284. PLoS One 2013;8:e76352.
31. Otto ME, Belohlavek M, Romero-Corral A, et al. Compari- 43. Kim D, Shim CY, Cho YJ, et al. Continuous Positive Air-
son of Cardiac Structural and Functional Changes in Obese way Pressure Therapy Restores Cardiac Mechanical Func-
Otherwise Healthy Adults With Versus Without Obstructive tion in Patients With Severe Obstructive Sleep Apnea: A
Sleep Apnea. Am J Cardiol 2007;99:1298-1302. Randomized, Sham-Controlled Study. J Am Soc Echocar-
32. Tugcu A, Guzel D, Yildirimturk O, Aytekin S. Evaluation of diogr 2019;32:826-835.
Right Ventricular Systolic and Diastolic Function in Patients 44. Kovács A, Lakatos B, Tokodi M, Merkely B. Right ven-
with Newly Diagnosed Obstructive Sleep Apnea Syndrome tricular mechanical pattern in health and disease: beyond
without Hypertension. Cardiology 2009;113:184-192. longitudinal shortening. Heart Fail Rev 2019;24:511-520.
33. Mendoza-Vázquez J, Steiner S, Esquinas AM. Acute and 45. Jasaityte R, Heyde B, D’hooge J. Current State of Three-
chronic effects of noninvasive ventilation on left and right Dimensional Myocardial Strain Estimation Using Echocar-
myocardial function in patients with obstructive sleep ap- diography. J Am Soc Echocardiogr 2013;26:15-28.
nea syndrome: a speckle tracking echocardiographic study. 46. Atsumi A, Seo Y, Ishizu T, et al. Right Ventricular Deforma-
Echocardiography 2016;33:1623-1624. tion Analyses Using a Three-Dimensional Speckle-Track-
34. Mandoli GE, De Carli G, Pastore MC, et al. Right cardiac ing Echocardiographic System Specialized for the Right
involvement in lung diseases: a multimodality approach Ventricle. J Am Soc Echocardiogr 2016;29:402-411.e2.
from diagnosis to prognostication. J Intern Med 2020 Sep 47. Smith BC, Dobson G, Dawson D, Charalampopoulos
30. doi:10.1111/joim.13179. A, Grapsa J, Nihoyannopoulos P. Three-dimensional
35. Trivedi SJ, Terluk AD, Kritharides L, et al. Right ventricu- speckle tracking of the right ventricle. J Am Coll Cardiol
lar speckle tracking strain echocardiography in patients 2014;64:41-51.
with acute pulmonary embolism. Int J Cardiovasc Imaging 48. Liu BY, Wu WC, Zeng QX, et al. Comparison of the capabil-
2020;36:865-872. ity of risk stratification evaluation between two- and three-
36. Li Y, Li H, Zhu S, et al. Prognostic Value of Right Ven- dimensional speckle-tracking strain in pre-capillary pulmo-
tricular Longitudinal Strain in Patients With COVID-19. nary hypertension. Pulm Circ 2019;9:2045894019894525.
JACC Cardiovasc Imaging 2020 Apr 28. doi:10.1016/j. 49. Huntgeburth M, Germund I, Geerdink LM, Sreeram N,
jcmg.2020.04.014. Udink Ten Cate FEA. Emerging clinical applications of
37. Altekin RE, Karakas MS, Yanikoglu A, et al. Determination strain imaging and three-dimensional echocardiography
of right ventricular dysfunction using the speckle tracking for the assessment of ventricular function in adult con-
echocardiography method in patients with obstructive sleep genital heart disease. Cardiovasc Diagn Ther 2019;9 (Suppl
apnea. Cardiol J 2012;19:130-139. 2):S326-S345.
Pictorial essay Med Ultrason 2021, Vol. 23, no. 2, 220-225
DOI: 10.11152/mu-2580
11st
Department of Obstetrics and Gynecology, University of Medicine and Pharmacy “Iuliu Hațieganu” Cluj-Napoca,
2Department of Obstetrics and Gynecology, University of Medicine, Pharmacy, Sciences and Technology, Târgu Mureș,
3Department of Obstetrics and Gynecology, University of Medicine and Pharmacy, Craiova, Romania
Abstract
Classically fetal heart evaluation is mandatory in the second trimester of pregnancy. Recent data suggest that the diagnosis
of congenital heart disease is feasible in the first trimester (FT) of pregnancy, especially for trained examiners. In this pictorial
essay we aim to illustrate in detail the particularity of the FT heart evaluation: the specific ultrasound techniques including
practical tips for the basic and extended first trimester fetal heart evaluation protocol. The diagnosis is possible by the use of
a variety of ultrasound techniques (B mode, CFM, Bidirectional Doppler, STIC, TUI) using the transabdominal/transvaginal
probes. An abnormal FT heart scan requires second trimester reassessment. Unfortunately, not all cardiac disease can be po-
tentially diagnosed in the first trimester. Based on their own experience the authors recommend that heart examination in the
FT should be attempted in all fetuses, even in low risk pregnancies, using at least the basic Doppler examination protocol.
Keywords: fetal heart; congenital heart defects; ultrasound; spatio-temporal image correlation (STIC); four chamber view
Fig 2. Examination of a normal fetal heart in the first trimester using the CFM transabdominal probe: a) four-chamber view – normal
diastolic flow from both right and left atrium into right and respectively left ventricle; b) “V” sign – aortic arch and pulmonary arch;
c) parasagittal view of the aortic and ductal arch
Fig 3. 3D technique for the evaluation of the fetal heart in the first trimester: a) spatiotemporal image correlation (STIC) acquisi-
tion with 3D rendered image – diastolic ventricular inflow and systolic aortic and pulmonary flow can be seen on the same image;
b) tomographic ultrasound imaging in a STIC acquisition.
222 Ioana Cristina Rotar et al First trimester fetal heart evaluation. A pictorial essay
Fig 5. Dextrocardia CFM mode: a) upper abdomen section – stomach on the lef, umbilical vein (blue) in the middle of the image;
b) four-chamber view – normal diastolic flow in atria and ventricles with the apex of the heart pointinig to the right; c) outflow tracts;
d) 3D rendering acquisition of the outflow tracts
let VSD. VSD in the first trimester are difficult to detect severe stenosis of the mitral and/or aortic valve or aor-
due to the small dimensions of the defect [12]. They are tic coarctation with no/little communication between the
rarely isolated, more frequently being encountered in the left ventricle and the aorta [12]. Less frequent the right
case of complex cardiac malformations.
Atrioventricular septal defect (AVSD) is a CHD de-
termined by an insufficient septation of the heart associ-
ated with anomalies of the AV valves. In figure 7 a com-
plete form of AV valve is presented: single ventricular
inflow though the unique atrioventricular valve towards
both ventricles. The genetic evaluation of fetuses with
AVSD is mandatory; an association with Down syn-
drome or trisomy 18 is frequently encountered especially
when a high nuchal translucency is detected.
An early morphologic evaluation of the ventricles in
the FT is possible. In figure 8a a ventricular dispropor-
tion can be seen in 4CV: a smaller left ventricle com-
pared to a normal right ventricle. In this case, the sec-
ond trimester follow-up revealed an unbalanced partial
atrioventricular septal defect with aortic arch hypoplasia. Fig 7. Atrioventricular septal defect: spatiotemporal image
Usually the hypoplastic left heart syndrome is due to the correlation acquisition CFM diastolic flow, in four-chamber
underdevelopment of the left ventricle secondary to the view and rendered image of the flow
sel view (3VV) only two vessels are seen: superior vena
cava and the aorta, the pulmonary artery not being visible
nor in B mode (fig 9a), nor in CFM (fig 9b). The 3VV
was essential for the diagnosis.
Aberrant right subclavian artery (ARSA) represent
an anatomic variant when the subclavian artery arises di-
rectly from the aortic arch instead of the brachiocephalic
trunk, more frequently encountered in Down syndrome
compared to euploid fetuses (fig 10) [12].
Color Doppler allows a good evaluation of the ven-
Fig 9. Abnormal three-vessel view in a fetus with complex con- triculoarterial connections, essential for the diagnosis of
genital heart defects with hypoplastic ductal arch: a) B mode; transposition of great vessel (fig 11).
b) CFM. SVC, superior vena cava; DescAo, descending aorta;
Ao arch, aortic arch. The performance of ultrasound scan between
11-14 WA
ventricle can be hypoplastic in the case of tricuspid atre-
sia (fig 8b) or pulmonary stenosis. In the B mode at the The visualization rates of fetal cardiac structures are
4CV the right ventricle is obviously smaller than the left higher towards the end of FT. Accordingly, CHD detec-
ventricle. tion rates varies from 20% at 11 WG up to 92% at 13
The assessment of the great vessels is an essential WG [13].
step. Anomalies of the great vessels can potentially be In Quarello et al study, the visualization rate of 4CV
diagnosed in the FT of pregnancy. In figure 9, a complex was 86% but 3VT was evaluated only in 79% of fetuses
case of CHD is demonstrated – a fetus with severe tri- [8]. The visualization rates of 4CV was not influenced
cuspid atresia and small right ventricle. In the three-ves- by the gestational age, but 3VT view was better seen
Fig 10. Aberrant right subclavian artery: a) retrotracheal route of the right subclavian artery; b) orthogonal acquisition with 3D
rendering
Fig 11. Transposition of the great vessels: a) aorta and pulmonary artery have a parallel route at the emergence from the ventricles.
There is not the normal crossing; b) 3D acquisition and rendered image shows that the artery that arise from the left ventricle has a
typical pulmonary bifurcation, confirming the diagnostic of transposition.
Med Ultrason 2021; 23(2): 220-225 225
when the CRL was >75 mm [8]. The use of 4CV section 4. Bishop KC, Kuller JA, Boyd BK, Rhee EH, Miller S, Bark-
alone had a sensitivity of only 45.71% compared to the er P. Ultrasound examination of the fetal heart. Obstet Gy-
4CV+3VT - 88.57% for the detection of CHD [7]. necol Surv 2017;72:54–61.
5. Li YF, Zhou KY, Fang J, Wang C, Hua YM, Mu DZ. Ef-
The use of transvaginal probes provides a higher
ficacy of prenatal diagnosis of major congenital heart
resolution particularly suitable for obese women or in
disease on perinatal management and perioperative mor-
the case of a retroverted uterus. Indeed, in 58%-62% of tality: a meta-analysis. World J Pediatr 2016;12:298-
cases the use of a 5-9 MHz TV probe obtains a satis- 307.
factory assessment of the FH structures, the visualization 6. García Fernández S, Arenas Ramirez J, Otero Chouza MT,
rates increasing linearly with the gestational age [14]. In Rodriguez-Vijande Alonso B, Llaneza Coto AP. Early fe-
a cross-sectional study that compared TA versus TV US, tal ultrasound screening for major congenital heart de-
the latter was usually superior before 14 GWs [14]. fects without Doppler. Eur J Obstet Gynecol Reprod Biol
The early detection rates of major CHD depend not 2019;233:93–97.
only on the protocol used, but also on the examiner skills; 7. Wiechec M, Knafel A, Nocun A. Prenatal detection of
ranging in an unselected population between 2.3-56% congenital heart defects at the 11- to 13-week scan using
a simple color Doppler protocol including the 4-cham-
[15]. In contrast, in high-risk populations, cardiac exam-
ber and 3-vessel and trachea views. J Ultrasound Med
ination reached a sensitivity of 78.5-90% with a 74.5% 2015;34:585–594.
in concordance with the second trimester diagnosis [15]. 8. Quarello E, Lafouge A, Fries N, Salomon LJ, CFEF. Basic
The use of STIC ameliorates the intra- and interob- heart examination: feasibility study of first-trimester sys-
server agreement [9,11]. tematic simplified fetal echocardiography. Ultrasound Ob-
stet Gynecol 2017;49:224–230.
Conclusions 9. Tudorache S, Cara M, Iliescu DG, Novac L, Cernea N. First
trimester two- and four-dimensional cardiac scan: intra-
The capacity to achieve good quality cardiac ultra- and interobserver agreement, comparison between methods
sound images, important for the confirmation of the nor- and benefits of color Doppler technique. Ultrasound Obstet
Gynecol 2013;42:659-668.
mality of fetal heart in the FT, is feasible and depends
10. Hernandez-Andrade E, Patwardhan M, Cruz-Lemini M,
not only on ultrasound probes and software, but, more
Luewan S. Early Evaluation of the Fetal Heart. Fetal Diagn
important, on the sonographer’s experience, commitment Ther 2017;42:161-173.
and awareness. New techniques of ultrasound – Doppler 11. 11.Turan S, Turan OM, Ty-Torredes K, Harman CR, Bas-
and 3/4D methods – greatly improve the diagnostic capa- chat AA. Standardization of the first-trimester fetal cardiac
bilities. The heart is a developing organ; therefore, even examination using spatiotemporal image correlation with
after a successfully complete heart scan in the FT, the tomographic ultrasound and color Doppler imaging. Ultra-
heart should be reassessed in every trimester due the po- sound Obstet Gynecol 2009;33:652–656.
tential evolution of the cardiac condition. 12. Abuhamad AZ, Chaoui R. A Practical Guide to Fetal Echo-
cardiography. Normal and abnormal hearts. Third edition.
Conflict of interest: none LWW, 2015.
13. Rasiah SV, Publicover M, Ewer AK, Khan KS, Kilby
References MD, Zamora J. A systematic review of the accuracy of
first-trimester ultrasound examination for detecting ma-
1. European Commission. Prevalence charts and tables. Avail- jor congenital heart disease. Ultrasound Obstet Gynecol
able at: https://eu-rd-platform.jrc.ec.europa.eu/eurocat/ 2006;28:110–116.
eurocat-data/prevalence_en. Accessed on 12.04.2020. 14. Vimpelli T, Huhtala H, Acharya G. Fetal echocardiography
2. Donofrio MT, Moon-Grady AJ, Hornberger LK, et al. Di- during routine first-trimester screening: a feasibility study
agnosis and treatment of fetal cardiac disease: a scientific in an unselected population. Prenat Diagn 2006;26:475-
statement from the American Heart Association. Circula- 482.
tion 2014;129:2183–2242. 15. Clur SA, Bilardo CM. Early detection of fetal cardiac ab-
3. AIUM Practice Parameter for the Performance of Fetal normalities: how effective is it and how should we manage
Echocardiography. J Ultrasound Med 2020;39:E5-E16. these patients? Prenat Diagn 2014;34:1235-1245.
Continuing education Med Ultrason 2021, Vol. 23, no. 2, 226-230
DOI: 10.11152/mu-2873
Abstract
Sprains are the most common injury of the ankle joint and the most common traumatic injury of the musculoskeletal
system. Ultrasound (US) examination of the posterior ankle joint is a challenge for the examiner. This paper focuses on this
difficult area and provides guidance on how to effectively perform US examination of the posterior ankle.
Keywords: ultrasound; posterior talofibular ligament; posterior tibiotalar ligament; posterior talocalcaneal ligament; flexor
hallucis longus tendon
Fig 2. a) US of the distal part of the Achilles tendon, using Fig 3. US of the middle part of the Achilles tendon, using a
a longitudinal scan; b) patient and probe positioning. ACH – longitudinal probe position. Star – posterior talocalcaneal liga-
Achilles tendon; K – Kager fad pad; star – retrocalcaneal bursa. ment; arrow – posterior tibiotalar ligament.
The posterior process of the talus has two tubercles: US of the posterior ankle joint
the medial and the lateral, and accurate imaging of the tu- There is uncertainty regarding the most convenient
bercles is crucial to visualization of the ligaments of the position to perform US of the posterior ankle. The pos-
posterior ankle joint. The posterior talofibular ligament terior ankle is most accessible when the patient is laying
attaches to the lateral tubercle, and the posterior part of in a prone position with the knee slightly bent and the
the deltoid ligament attaches to the medial tubercle. As ankle joint held at an angle of 90º with the weight of the
with most joints, the ligaments provide direct reinforce- foot resting on the toes. However, in the case of recent
ment of the joint capsule and form the deepest layer of injury, this position may cause considerable pain, so it is
soft tissue, located adjacent to the articular surface of the often necessary to adapt to the situation and instead ex-
bones forming the joint. Another important anatomical amine the patient in a supine position with the knee bent
structure, the flexor hallucis longus tendon, is located to approximately 90º and the foot resting on the floor.
in the groove between the two tubercles of the posterior This alternative position is often sufficient to adequately
process of the talus, and is important to visualize in the evaluate the ligaments.
context of diagnostic imaging following ankle injury. Su- In the first part of the examination, the transducer is
perficial to the joint capsule and ligamentous structures is placed along the long axis of the limb over the central
the connective tissue of the Kager fad pad and more su- part of the Achilles tendon. One should remember at the
perficial to this is the Achilles tendon. The retrocalcaneal beginning of the examination to set an appropriate depth
bursa is located between the distal part of the Achilles and focus and to choose an appropriate frequency. A mul-
tendon, the upper surface of the calcaneus and the con- ti-frequency probe with a range of 3–12 MHz, with the
nective tissue of the Kager fad pad (fig 2). frequency set to the upper range is recommended. The
goal is to visualize the outline of the bony structures and
Fig 4. a) Posterior talofibular ligament (PTFL), longitudinal Fig 5. Posterior tibiotalar ligament (part of the deltoid liga-
view and b) probe position. MT – medial tubercle of the pos- ment), longitudinal scan; b) patient and probe positioning.
terior process of the talus; LT – lateral tubercle of the posterior MM – malleolus mediualis; MT – medial tubercle of the pos-
process of the talus; star - groove between the two tubercles of terior process of the talus; LT – lateral tubercle of the posterior
the posterior process of the talus (flexor hallucis longus tendon process of the talus; star – flexor hallucis longus tendon; arrows
space); arrows – posterior talofibular ligament (PTFL). – posterior tibiotallar ligament.
228 Wojciech Konarski et al The utility of ultrasound in the diagnostic evaluation of the posterior ankle joint
Fig 6. a) Flexor hallucis longus (FHL), transverse scan; b) pa- Fig 7. a) Flexor hallucis longus (FHL), longitudinal section;
tient and probe positioning. MT – medial tubercle of the pos- b) patient and probe positioning.
terior process of the talus; LT – lateral tubercle of the posterior
process of the talus; star – flexor hallucis longus tendon.
Post-traumatic lesions of the flexor hallucis longus Fig 12. Flexor hallucis longus (FHL) tenosynovitis, transverse
tendon are rare (fig 11). The most frequently observed scan (a) due to the presence of os trigonum (arrow) (b).
pathology is tenosynovitis (fig 12), often associated with
posterior ankle impingement syndrome, which is caused 4. Fursdon T, Platt S. The incidence and significance of pos-
by the presence of a large os trigonum bone. The medial terior talofibular ligament injury on magnetic resonance im-
surface of the bone may adhere to the tendon sheath and aging. Orthop Proc 2018;94B SUPP_XXII:56.
5. Kumai T, Takakura Y, Rufai A, Milz S, Benjamin M. The
as a result of mechanical irritation, cause exudative ten-
functional anatomy of the human anterior talofibular liga-
osynovitis This condition most often occurs in athletes ment in relation to ankle sprains. J Anat 2002;200:457-
such as footballers or ballet dancers and is associated 465.
with frequent flexion of the ankle [11]. 6. Arthurs G, Nicholls B. Ultrasound in anesthesia, critical
care and pain management. Cambridge University Press
Conclusion 2016.
7. Park JW, Lee SJ, Choo HJ, Kim SK, Gwak HC, Lee
US is a valuable modality to assess the structures
SM. Ultrasonography of the ankle joint. Ultrasonography
within the posterior ankle joint. The main utility is in the 2017;36:321-335.
diagnostic imaging of the posterior talofibular ligament 8. Fessell DP, Vanderschueren GM, Jacobson JA, et al. US of
and the posterior portion of the deltoid ligament follow- the ankle: technique, anatomy, and diagnosis of pathologic
ing ankle injury. US is also useful in the evaluation of conditions. Radiographics 1998;18:325-340.
posterior ankle impingement syndrome caused by the 9. George J, Jaafar Z, Hairi IR, Hussein KH. The correlation
presence of a hypertrophied os trigonum bone, which can between dynamic ultrasound evaluation and clinical laxity
cause an exudative flexor hallucis longus tenosynovitis. grading of ATFL and CFL tears among athletes. J Sports
US also provides relatively low cost diagnostic imaging Med Phys Fitness 2020;60:749-757.
following ankle trauma in real time. 10. Margetic P, Salaj M, Lubina IZ. The Value of Ultrasound
in Acute Ankle Injury: Comparison With MR. Eur J Trauma
References Emerg Surg 2009;35:141-146.
11. Yasui Y, Hannon CP, Hurley E, Kennedy JG. Posterior
1. Polzer H, Kanz KG, Prall WC, et al. Diagnosis and treat- ankle impingement syndrome: A systematic four-stage ap-
ment of acute ankle injuries: development of an evidence- proach. World J Orthop 2016;7:657-663.
based algorithm. Orthop Rev (Pavia) 2012;4:e5. 12. Bianchi S, Martinoli C, Gaignot C, De Gautard R, Meyer
2. Boruta PM, Bishop JO, Braly WG, Tullos HS. Acute Lat- JM. Ultrasound of the Ankle: Anatomy of the Tendons,
eral Ankle Ligament Injuries: A Literature Review. Foot Bursae, and Ligaments. Semin Musculoskelet Radiol
Ankle 1990;11:107-113. 2005;9:243-259.
3. Fong DT, Man CY, Yung PS, Cheung SY, Chan KM. Sport- 13. Khoury V, Guillin R, Dhanju J, Cardinal É. Ultrasound of
related ankle injuries attending an accident and emergency Ankle and Foot: Overuse and Sports Injuries. Semin Mus-
department. Injury 2008;39:1222-1227. culoskelet Radiol 2007;11:149-161.
Case report Med Ultrason 2021, Vol. 23, no. 2, 231-234
DOI: 10.11152/mu-2325
1“Niculae Stancioiu” Heart Institute, 2“Iuliu Haţieganu” University of Medicine and Pharmacy, 3Clinical Rehabilitation
Hospital, Cardiology Department, Cluj-Napoca, România
Abstract
Anomalous left coronary artery from the pulmonary artery (ALCAPA) syndrome is a rare congenital coronary anomaly,
which can cause potentially fatal complications, such as heart failure, myocardial infarction and sudden cardiac death. Only a
few patients left untreated survive to adulthood. We highlight the importance of multimodal imaging in the diagnosis of AL-
CAPA syndrome in a young asymptomatic female patient with inducible ischemia on exercise. The patient was successfully
treated with surgery.
Keywords: multimodal imaging; ALCAPA; coronary angiography; echocardiography
Discussion
Fig 3. CCTA revealing an anomalous origin of the LCA from the PA with a normal origin of the RCA(A) and many collaterals (red
arrows) between LAD and RCA (B).
grade flow from the LCA to the PA and the collaterals the aortic buds. On the other hand, LVNC is caused by
with systolic and diastolic blood flow [5]. Moreover, the arrest of the embryogenesis of the endocardium and
speckle tracking echocardiography reveals reduced lon- myocardium, with coronary circulation being developed
gitudinal and circumferential strain in the regions corre- simultaneously during this process, when intratrabecular
sponding to the LCA [2]. CCTA allows excellent spatial recesses are reduced to capillaries [2]. In our patient, both
resolution to establish the origin and course of the CA [2]. echocardiography and MRI showed a thickened lateral
MRI can be useful due to the benefit of LGE, which in- LV wall with positive criteria of LVNC, but with multiple
dicates fibrosis secondary to chronic ischemia [6]. Some milimetric lodges inside the myocardium, demonstrating
of these findings were also present in our patient. There the connection between the two processes (development
were septal and lateral color flow signals on echocardi- of the CA and compactation of the LV).
ography, with reduced regional strain at the level of the If untreated, ALCAPA syndrome has a high mortality
lateral LV wall. The CCTA revealed the abnormal course (80-90%) [2]. SCD occurs mainly in young athletes and
of the RCA with anastomoses between the right and left basketball players [1]. A literature review of 151 patients
CA. MRI demonstrated fibrosis at the level of the lateral with this pathology found that 14% were asymptomatic
and anterior septal walls as a sign of chronic ischemia. and 62% of those with SCD were asymptomatic before
The connection between ALCAPA syndrome and the diagnosis was established [7].
LVNC can be explained by early embryonic develop- The particularity of our case consists in the rarity of
ment. ALCAPA is the result of abnormal septation of the survival in patients with ALCAPA syndrome when left
conotruncus into the aorta and pulmonary artery or due untreated, moreover with the patient being asymptomat-
to persistence of the pulmonary buds and involution of ic and physically active. Even though ALCAPA is rare
in the adult population, the diagnosis is essential since
early treatment may prevent myocardial damage. Cur-
rent guidelines indicate that such patients may return to
competitive sports 3 months postoperatively, provided
that they remain asymptomatic and an exercise stress test
does not show ischemia or important arrhythmias [3].
In conclusion we underline the importance of im-
aging in the early diagnosis of patients with ALCAPA
syndrome. Furthermore, we highlight the connection be-
tween CA anomalies and LVNC, as these two processes
coincide during early embryogenesis.
References
1. Harmon KG, Asif IM, Maleszewski JJ, et al. Incidence,
cause, and comparative frequency of sudden cardiac death
Fig 4. Three-months postoperative TEE showing significant in national collegiate athletic association athletes: a decade
reduction of the collateral vessels. in review. Circulation 2015;132:10-19.
234 Raluca Tomoaia et al The role of multimodal imaging in the diagnosis of an asymptomatic patient with congenital anomaly
2. Elumalai G, Sujitha AS. “Anomalies origin of left coronary and outcomes of anomalous left coronary artery from the
artery” its embryological basis and clinical significance. pulmonary artery. J Am Soc Echocardiogr 2017;30:896-
Elixir Embryology 2016;100:43446-43449. 903.
3. Brothers JA, Frommelt MA, Jaquiss RDB, Myerburg RJ, 6. Boutsikou M, Shore D, Li W, et al. Anomalous left coro-
Fraser CD, Tweddell JS. Expert Consesnsus guidelines: nary artery from the pulmonary artery (ALCAPA) diag-
Anomalous aortic origin of a coronary artery. J Thorac Car- nosed in adulthood: Varied clinical presentation, thera-
diovasc Surg 2017;153:1440-1457. peutic approach and outcome. Int J Cardiol 2018;261:
4. Dilawar M, Ahmad Z. Anomalous left coronary artery from 49-53.
pulmonary artery: Case series and brief review. OJPed 7. Memon MKY, Amanullah M, Atiq M. Anomalous Left
2012;2:77-81. Coronary Artery from Pulmonary Artery: An Important
5. Patel SG, Frommelt MA, Frommelt PC, Kutty S, Cramer Cause of Ischemic Mitral Regurgitation in Children. Cureus
JW. Echocardiographic diagnosis, surgical treatment, 2019;11:e4441.
Case report Med Ultrason 2021, Vol. 23, no. 2, 235-237
DOI: 10.11152/mu-2503
¹Department of Gastroenterology and Hepatology, “Victor Babeș ”University of Medicine and Pharmacy, ²Neuromed
Diagnostic Imaging Center Timișoara, Romania
Abstract
Primary biliary tract neuroendocrine tumors (NETs) are extremely rare tumors that account for 0.2-2% of all gastrointes-
tinal neuroendocrine tumors. The typical presentation is with jaundice and other symptoms related to biliary obstruction.We
present a case of right hepatic duct NET in a 27-year-old female patient, asymptomatic, presented for a routine ultrasound
examination that revealed moderate dilatation of the intrahepatic biliary ducts and a 20 mm hyperechoic lesion in the right
hepatic biliary duct. Additional imaging was performed with the presumptive diagnosis of cholangiocarcinoma. After surgery,
the histopathological and immunohistochemical report was conclusive for the diagnosis of G2 well-differentiated NET.
Keywords: neuroendocrine tumor; right hepatic bile duct; biliary obstruction
Discussion
Fig 2. CEUS examination: a) homogenous hyperenhancement pattern in the arterial phase; b) mild washout starting in the portal
vascular phase; c) obvious washout in the late vascular phase.
Med Ultrason 2021; 23(2): 235-237 237
evidence for its origin in metaplastic endocrine cells. Am J
Gastroenterol1991;86:1073-1076.
5. Noronha YS, Raza AS. Well-differentiated neuroendocrine
(carcinoid) tumors of the extrahepatic biliary ducts. Arch
Pathol Lab Med 2010;134:1075-1079.
6. Kuwabara H, Uda H. Small cell carcinoma of the gall-
bladder with intestinalmetaplastic epithelium. Pathol Int
1998;48:303-306.
7. Bosman FT, Carneiro F, Hruban RH, Theise ND. WHO
Classification of Tumours of the Digestive System. 4th ed.
Lyon: IARC, 2010.
8. Lloyd RV, Osamura RY, Klöppel G, Rosai J. WHO Clas-
Fig 3. T1- Gadolinium postcontrast sequences, late phase: cor- sification of Tumours of Endocrine Organs. 4th ed. Lyon:
onal (a) and axial plane (b). IARC, 2017
9. UedaY, Toyama H, Fukumoto T, Ku Y. Prognosis of Pa-
tients with Neuroendocrine Neoplasms of the Pancreas ac-
The literature search did not reveal other cases of
cording to the World Health Organization 2017 Classifica-
NET isolated in the right hepatic biliary duct, possibly tion. JOP 2017;18 S(3):366-370.
making our case the first reported case of an intraductal 10. Michalopoulos N, Papavramidis TS, Karayannopoulou G,
NET of the right hepatic biliary duct. Pliakos I, Papavramidis ST, Kanellos I. Neuroendocrine
In conclusion, biliary tract NETs have as a typical tumors of extrahepatic biliary tract. Pathol Oncol Res
presentation, jaundice and other symptoms related to bil- 2014;20:765-775.
iary obstruction, but patients can also be asymptomatic. 11. Bhandarwar AH, Shaikh TA, Borisa AD, Palep JH, Patil
Due to the absence of specific symptoms, a correct pre- AS, Manke AA. Primary neuroendocrine tumor of the left
operative diagnosis is rare: diagnosis is usually made hepatic duct: a case report with review of the literature.
postoperatively, based on the histopathology exam. Radi- Case Rep Surg 2012;2012:786432.
12. Nesi G, Lombardi A, Baltignani G, Ficari F, Ru-
cal tumor surgery is the only available curative approach,
bio CA, Tonelli F. Well-differentiated endocrine tu-
with high survival rates.
mor of the distal common bile duct: a case study
References and literature review. Virchows Arch 2006;449:104-
111.
1. Khan FA, Stevens-Chase A, Chaudhry R, Hashmi A, Edel- 13. Oskuie AE, Valizadeh N. Carcinoid tumor of the common
man D, Weaver D. Extrahepatic biliary obstruction second- bile duct misdiagnosed as cholangiocarcinoma. Middle
ary to neuroendocrine tumor of the common hepatic duct. East J Cancer 2011;2:139-142.
Int J Surg Case Rep 2017;30:46-49. 14. Costin AI, Păun I, Păun M, Constantin VD, Vârcuş F. Pri-
2. Modlin IM, Sandor A. An analysis of 8305 cases of Carci- mary neuroendocrine tumors–an extremely rare causeof
noid tumors. Cancer 1997;79:813-829. obstruction of extrahepatic bile ducts: a case report. Rom J
3. Lauffer JM., Zhang T, Modlin IM. Review article: current Morphol Embryol 2017;58:641-644.
status of gastrointestinal carcinoids. Aliment Pharmacol 15. Choi J, Lee KJ, Kim SH, Cho MY. Preoperative diagnosis
Therap 1999;13:271-287. of well-differentiated neuroendocrine tumor in common
4. Barron-Rodriguez LP, Manivel JC, Mendez-Sanchez N, hepatic duct by brush cytology: A case report. Diagn Cyto-
Jessurun J. Carcinoid tumors of the common bile duct: pathol 2019;47:720-724.
Letter to the Editor Med Ultrason 2021, Vol. 23, no. 2, 238-247
Koya Akutagawa
To the Editor,
1Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital, Bei-Hu Branch and
National Taiwan University College of Medicine, Taipei, Taiwan, 2Center for Regional Anesthesia and Pain Medicine,
Wang-Fang Hospital, Taipei, Medical University, Taipei, Taiwan, 3Physical Medicine and Rehabilitation, Taipei
Veterans General Hospital, Taipei, Taiwan, 4Department of Physical and Rehabilitation Medicine, Hacettepe University
Medical School, Ankara, Turkey
To the Editor
muscle strength and more than 50% decrease in thumb 106-2314-B-002-180-MY3 and 109-2314-B-002-114-
pain. MY3), and Taiwan Society of Ultrasound in Medicine.
The RBMN provides innervation to the thenar mus-
cles [2]. It mostly emerges from the main trunk of the me- References
dian nerve distal to the carpal tunnel outlet and ascends
1. Chen IJ, Chang KV, Lou YM, Wu WT, Özçakar L. Can ul-
vertically to the palmar surface through the anterior edge
trasound imaging be used for the diagnosis of carpal tunnel
of the transverse carpal ligament [3]. The RBMN further syndrome in diabetic patients? A systemic review and net-
curves backward to pierce and innervate the thenar mus- work meta-analysis. J Neurol 2020;267:1887-1895.
cles (fig 1D). The RBMN neuropathy is uncommon and 2. Riegler G, Pivec C, Platzgummer H, et al. High-resolution
can result from neurogenic tumours, compression of the ultrasound visualization of the recurrent motor branch of
thenar muscles during long-distance cycling and iatro- the median nerve: normal and first pathological findings.
genic injury. In this case, we speculated that the anteced- Eur Radiol 2017;27:2941-2949.
ent trigger point injection might be the culprit of RBMN 3. Smith J, Barnes DE, Barnes KJ, et al. Sonographic visual-
neuropathy. Moreover, the subsequent thenar fibrosis ization of thenar motor branch of the median nerve: A ca-
could have led to entrapment of the nerve across its pas- daveric validation study. PM R 2017;9:159-169.
4. Lin CP, Chang KV, Huang YK, Wu WT, Özçakar L. Regen-
sage through the palmar fascia. Therefore, the US-guided
erative injections including 5% dextrose and platelet-rich
perineural hydro-dissection relieved the focal compres-
plasma for the treatment of carpal tunnel syndrome: A sys-
sion and facilitated the neural recovery [4,5]. tematic review and network meta-analysis. Pharmaceuti-
cals (Basel) 2020;13:49.
Acknowledgement: The current research project was 5. Chang KV, Wu WT, Özçakar L. Ultrasound imaging and
supported by National Taiwan University Hospital, Bei- guidance in peripheral nerve entrapment: hydrodissection
Hu Branch, Ministry of Science and Technology (MOST highlighted. Pain Manag 2020;10:97-106.
Department of Physical Medicine and Rehabilitation, National Taiwan University Hospital and National Taiwan
University College of Medicine, Taipei, Taiwan
To the Editor, took a walk. She reported no falling episode at the time.
Initially, she could walk with a walker but, during the
A 72-year-old woman with history of rectal ad- following days, her hip pain progressed and finally she
enocarcinoma and brain metastasis (but no documented became wheelchair-bound.
bone metastasis) had in the recent medical history left On examination, she was intolerant to both the pas-
proximal femoral vein thrombosis treated with an anti- sive and active range of motion test due to severe right
coagulation agent. One month before visiting the clinic hip pain. The pain worsened on weight bearing and im-
she had sudden onset of a severe right hip pain when she proved when resting supine. Ultrasound evaluation dem-
onstrated discontinued cortex of the right femoral neck
and a hypervascular hyperechoic amorphous soft tissue
Received 30.03.2021 Accepted 18.04.2021 was noted just over the cortical cleft (fig 1a, b). These
Med Ultrason
findings were consistent with right hip fracture at the
2021, Vol. 23, No 2, 240-241, DOI: 10.11152/mu-3180,
Corresponding author: Dr. Shaw-Gang Shyu femoral neck with callus formation. The hip plain film
Department of Physical Medicine and disclosed displaced hip fracture of the right femoral neck,
Rehabilitation, National Taiwan University Garden type IV (fig 1c). Therefore, she underwent bipo-
Hospital, 7 Zhongshan South Road,
Zhongzheng District, 100 Taipei City, Taiwan
lar hemiarthroplasty of the right hip. Pathological exam
Phone: 886-2-23123456-67587 disclosed osteoporosis with marrow atrophy and no evi-
E-mail: taotaohughs@gmail.com dence of malignancy.
Med Ultrason 2021; 23(2): 238-247 241
overall mortality rate of conservative treatment is four
times as high in one year [1]. Decision of the surgical
method depends on the likelihood of blood supply resto-
ration. Internal fixation is indicated in the femoral neck
fracture with age less than 60 or non-displaced fracture,
while arthroplasty is indicated in displaced fracture for
those aged more than 60 [2].
Ultrasound is a convenient image modality that al-
lows early diagnosis of the hip fracture, especially when
occult fracture is undetectable in the x-ray. A previous
study disclosed that compared with magnetic resonance
imaging, sensitivity and specificity of ultrasound for hip
fracture was 100% and 65% respectively [3]. Classical
ultrasound findings include fracture line with cortical dis-
continuity, peritrochanteric hypoechoic hematoma or flu-
id collection, soft tissue swelling and callus formation [3].
This case highlights the importance of the detection
of a potential hip fracture in a patient with negative high
energy trauma or fall history. In view of the high cost-
Fig 1. Coronal view of the right femoral ultrasound shows dis-
continued bony cortex with penetrating ultrasound beam (white effectiveness and accessibility, we suggest the physician
arrow) at the right femoral neck. A hyperechoic amorphous soft should become familiar with the clinical picture and ul-
tissue (black arrow) is also noted; b) Doppler mode showed trasound findings of the hip fracture.
hypervascularity of the amorphous soft tissue; c) the right
hip x-ray reveals right femoral neck fracture, Garden type IV References
(white arrowhead).
1. Tay E. Hip fractures in the elderly: operative versus non-
Hip fracture is a debilitating health issue that usually
operative management. Singapore Med J 2016;57:178-181.
results in a decreased quality of life and marked mobil- 2. Bhandari M, Swiontkowski M. Management of acute hip
ity and mortality. It deserves special attention when the fracture. N Engl J Med 2017;377:2053-2062.
patient has disabling hip pain or the problem of weight 3. Safran O, Goldman V, Applbaum Y, et al., Posttraumatic
bearing. Although the patient usually has a functional de- painful hip: sonography as a screening test for occult hip
cline after a fracture reduction and fixation operation, the fractures. J Ultrasound Med 2009;28:1447-1452.
When meniscus ‘tears’ make the Baker’s cyst ‘cry’: a story on knee
ultrasound
Carmelo Pirri1, Carla Stecco2, Nina Pirri3, Raffaele De Caro2, Levent Özçakar4
1Department of Neurosciences, Institute of Human Anatomy, University of Padova, Padova, Italy, 2School of Medicine
and Surgery, University of Messina, Messina, Italy, 3Department of Physical and Rehabilitation Medicine, Hacettepe
University Medical School, Ankara, Turkey
To the Editor,
Received 28.12.2020 Accepted 03.01.2021
Med Ultrason
2021, Vol. 23, No 2, 241-242, DOI: 10.11152/mu-3193, A 60-year-old female patient was seen because of
Corresponding author: Carmelo Pirri pain and functional limitation in the right knee for the last
Department of Neurosciences, Institute of
two months. The pain was worse at night and when run-
Human Anatomy, University of Padova,
Via Gabelli 67, 35121, Padova, Italy ning. She declared that she had medial meniscus and an-
E-mail: carmelop87@hotmail.it terior cruciate ligament tears for about five years and that
242 Carmelo Pirri et al When meniscus ‘tears’ make the Baker’s cyst ‘cry’: a story on knee ultrasound
1Department of General Surgery, “Dunarea de Jos” University of Galati, Faculty of Medicine and Pharmacy, Galati,
Romania, 2Service de Chirurgie Digestive, Unité 23, Chirurgie viscérale, Centre Hospitalier de Mâcon, France,
3Department of Morphological and Functional Sciences, “Dunarea de Jos” University of Galati, Faculty of Medicine
To the Editor,
1Department of Ultrasound, 2Department of Radiology, 3Department of Nephrology, The Affiliated Dongyang Hospi-
tal of Wenzhou Medical University, Dongyang, Zhejiang, China
Fig 1. The lesion is abutting posterior aspect of the left lobe of the thyroid and seems connected with the esophagus (arrow); b) No
apparent blood flow signal in or around the lesion; c) Ultrasound imaging reveals a gas-filled lumen projecting to the left lobe of the
thyroid; d) CT reveals a low-density lesion with a clear contour and partial gas-fluid filled (arrow); e) Barium swallow demonstrates
a barium-filled pouch projecting from the esophagus.
1Department of Ultrasound, 2Department of Radiology, 3Department of Nephrology, The Affiliated Dongyang Hospi-
tal of Wenzhou Medical University, Zhejiang, China
To the Editor, the blinded status of the index test was not definitely
reported. Furthermore, concerning the flow and timing
We read with great interest the manuscript of Ge et al domain, one study [4] should be considered as having an
[1]. We strongly agree with the authors about the impor- unknown risk as it did not definitely report the interval
tance of contrast-enhanced ultrasound in the diagnosis of time between the index test and the reference standard.
bladder cancer, but we would like to pay attention to sev- The authors found that, after the test, the positive
eral important missing aspects in the article. likelihood ratio increased from 20% to 70%, while the
We reveal some different results regarding the quality negative likelihood ratio decreased to 2%. We think the
assessment of the included studies in the article [1]. Con- interpretation is not appropriate. The rational interpreta-
cerning the reference standard domain, 3 studies [2-4] tion is that patients with a probability of 20% of develop-
should be considered as an unknown risk of bias because ing the disease show a probability of 70% of the disease
when a positive result of the test, while a probability of
Received 10.03.2021 Accepted 18.04.2021 2% of the disease when it is a negative result.
Med Ultrason
Ge et al [1] revealed that the pooled sensitivity
2021, Vol. 23, No 2, 245-246, DOI: 10.11152/mu-3146,
Corresponding author: Jiangfeng Wu (I2=62.02%, p=0.03>0.01) and specificity (I2=45.69%,
Department of Ultrasound, The Affiliated p=0.12>0.01) indicated no significant heterogeneity. But
Dongyang Hospital of Wenzhou Medical as described in the statistical analysis that if the outcome
University, 60 Wuning West Road,
Dongyang 322100, Zhejiang, China
was I2>0.5 or p<0.1 a random-effects model was selected
E-mail: wjfhospital@163.com to indicated the heterogeneity’s result. Therefore, the
Phone: 18257937213 p>0.01 indicating no significant heterogeneity was in-
246 Xin-Yue Ge et al Author’s response
References
1. Ge X, Lan ZK, Chen J, Zhu SY. Effectiveness of contrast- Fig 1. Forest plot of CEUS diagnosing bladder cancer, showing
enhanced ultrasound for detecting the staging and grading sensitivity and specificity
of bladder cancer: a systematic review and meta-analysis.
Med Ultrason 2021;18;23:29-35. 4. Li Q, Tang J, He E, Li Y, Zhou Y, Wang B. Differentiation
2. Caruso G, Salvaggio G, Campisi A, et al. Bladder tumor between high- and low-grade urothelial carcinomas using
staging: comparison of contrast-enhanced and gray-scale contrast enhanced ultrasound. Oncotarget 2017;8:70883-
ultrasound. AJR Am J Roentgenol 2010;194:151-156. 70889.
3. Drudi FM, Di Leo N, Malpassini F, Antonini F, Corongiu 5. Gupta VG, Kumar S, Singh SK, Lal A, Kakkar N. Con-
E, Iori F. CEUS in the differentiation between low and trast enhanced ultrasound in urothelial carcinoma of urinary
high-grade bladder carcinoma. J Ultrasound 2012;15:247- bladder: An underutilized staging and grading modality.
251. Cent European J Urol 2016;69:360-365.
Author’s response
1Department of Medical Ultrasound, First Affiliated Hospital of Guangxi Medical University, Nanning, 2Department
of Medical Ultrasound, Fourth Affiliated Hospital of Guangxi Medical University, Liuzhou, 3Department of Medical
Ultrasound, Liuzhou People’s Hospital affiliated to Guangxi Medical University, Liuzhou, Guangxi, China
Dear Editor, fore, 3 studies [1-3] were considered low risk. As for the
Li et al study [3], in another study by the same group [4]
We appreciate Wu et al’s interest in our review and published in 2012, the interval time between the index
we have read with interest the comments. test and the reference standard was mentioned and we
Before the literature search, we limited the inclusion took it in consideration.
criteria. The reference standards were considered to be Regarding probability, our interpretation is as fol-
cystoscopy and/or transurethral resection of bladder tu- lows: the pre-test probability is the probability of a blad-
mors and, concerning the index test domain, the readers der cancer (BC) being detected without taking contrast-
were blinded to the final diagnoses of the patients, there- enhanced ultrasound (CEUS) into account. The post-test
Med Ultrason 2021; 23(2): 238-247 247
probability takes into account the results of CEUS. When 2. Drudi FM, Di Leo N, Malpassini F, Antonini F, Corongiu E,
detection of BC was based on a CEUS-positive result, Iori F. CEUS in the differentiation between low and high-grade
there was a 70% “post-test” probability of detecting a bladder carcinoma. J Ultrasound 2012;24;15(4):247-251.
3. Li QY, Tang J, He EH, Li YM, Zhou Y, Wang BJ. Dif-
subsequent BC. With a negative CEUS, the “post-test”
ferentiation between high- and low-grade urothelial car-
probability of detecting BC dropped to 2%. Consistent
cinomas using contrast enhanced ultrasound. Oncotarget
with other diagnostic meta-analysis articles [5,6]. 2017;10;8(41):70883-70889.
We believe that moderate heterogeneity exists, so we 4. Li QY, Tang J, He EH, et al. Clinical utility of three-dimen-
stated in the discussion that the P-SEN and PLR− with sional contrast-enhanced ultrasound in the differentiation
95%CI forest plots displayed moderate heterogene- between noninvasive and invasive neoplasms of urinary
ity and we made corresponding explanations and analy- bladder. Eur J Radiol 2012; 81:2936-2942.
ses. 5. Sang L, Wang XM, Xu DY, Cai YF. Accuracy of shear
We agree with your valuable comments on the Gupta wave elastography for the diagnosis of prostate cancer: A
et al study [8]. We included this article in analysis be- meta-analysis. Sci Rep, 2017; 7(1): 1949.
cause in this study CEUS was a good alternative for pre- 6. Yang Y, Zhao XX, Shi JW, Huang Y. Value of shear wave
elastography for diagnosis of primary prostate cancer: a
operatively T staging and grading of BC.
systematic review and meta-analysis. Med Ultrason 2019,
21(4): 382-388.
References
7. Gupta VG, Kumar S, Singh SK, Lal A, Kakkar N. Con-
1. Caruso G, Salvaggio G, Campisi A, et al. Bladder tumor trast enhanced ultrasound in urothelial carcinoma of urinary
staging: comparison of contrast-enhanced and gray-scale bladder: An underutilized staging and grading modality.
ultrasound. AJR Am J Roentgenol 2010;194(1):151-156. Cent European J Urol 2016;69:360-365.
Guidelines for Authors
Medical Ultrasonography is the official publication of the b. Text and page layout
Romanian Society for Ultrasonography in Medicine and Biol- The papers submitted for publication must be write in
ogy (SRUMB). Microsoft Word program, in 12p Times New Roman font, 1.5
The journal aims to promote ultrasound diagnosis by line spacing. Page size should be A4, margins must be normal
publishing papers that deal with fundamental and clinical (2.54 mm in top, bottom, right and left). Page numbering must
research, scientific reviews, clinical case reports, progress in begin with the title page. No numbering of the lines is needed.
ultrasound physics or in the field of medical technology and c. Text length, abbreviations and measurement units
equipment, as well as educational papers, special reports and The length of the manuscripts: maximum 5000 words
letters to the editor. for original paper, 6000 for reviews, 2000 for pictorial essay,
The journal is published quarterly and papers are accepted 1500 for case report, 500 for letter to editor (including the
for publication in English language. abstract, keywords, references, legend and tables). For letter
The official title abbreviation is Med Ultrason. It should to editor only one imagine or table is accepted.
be used in citations, footnotes, legends for figures and all bib- The editors reserve the right of condensing any paper sub-
liographic references. mitted.
Use only standard abbreviations. Avoid abbreviations in
1. Copyright the title and abstract. Explain every abbreviation before to use
it.
Submitting a scientific paper to Medical Ultrasonography Measurements of length, height, weight, and volume
for publishing is subject to compliance with the following should be reported in metric units. All hematological and clin-
statements: ical chemistry measurements should be reported in the metric
• the paper is original and has not been published in other system in terms of the International System of Units (SI).
journals or books;
• the paper has not been sent or is not under consideration for d. Images and tables
publication elsewhere; Imagines (figures) and tables should be grouped in a dis-
• all authors agree upon publication of the paper. tinct section. They must be numbered according to the order in
The statements can be downloaded from the journal web- which they appear in the text- the imagines (figures) in Arabic
site. numeral (ex. fig 1) and the tables in Roman numeral (ex. tab I).
All these statements should be included in a formal dec- Do not send them in PDF format!
laration signed by all the authors. The filled in, signed and The captions for figures (images) must be typed on a sepa-
scanned images of these two forms have to be uploaded at the rate page entitled “Legend for figures”. Each table must have
last step of the submission process (Step 4. Uploading Sup- a title. Images or tables should not appear in the text; the de-
plementary Files). sired location for insertion should be indicated by means of a
In cases where the paper is accepted for publication, copy- paragraph, such as: (location for figure no….) or (location for
right will be transferred to Medical Ultrasonography and the table no…)
“Iuliu Hatieganu” Medical Publishing House. Authors must Only high quality images will be accepted for publica-
agree to undertake all responsibility for the scientific content tion. File formats: BMP or TIFF for images, 300 dpi, image
and originality of the paper; Medical Ultrasonography will width 8 cm (single column) or 16.7 cm (double column), for
take no responsibility whatsoever in this respect. color images color mode should be CMYK.
The tables (design: Table Grid) are accepted as word doc-
2. Preparing the manuscript uments included at the end of the main document (see below).
In cases where reproduction of previously published im-
2.1 General ages is intended, it is necessary to attach the written consent of
the author and of the publishing house where it was formerly
a. Style & language published. All prospective or experimental papers involving
Manuscripts should be prepared according to the style of human subjects or experimental animals must include the
the journal. They should be written in concise and grammati- agreement granted by the medical ethics commission of the
cally correct US English. Authors should ask for assistance if institution where the research was conducted.
not writing in their native language. If the manuscript reports medical research involving hu-
Papers not conforming to the style of the journal, incom- man subjects, authors must include a statement confirming
prehensible, written in inappropriate English, or not submit- that informed consent was obtained from all subjects, accord-
ing to the World Medical Association Declaration of Helsinki,
ted strictly according to the journal guidelines will be re-
revised in 2000, Edinburgh.
turned to the authors for revision, without peer reviewing.
If the manuscript is not complying to accepted standards 3. Structure of the submission files
of English usage, the authors may be required to bear the cost
of English supervision/ translation. The manuscript has to be submitted in this form:
a) in step 2 of submission process, the word document (the a) Article:
original file) that is uploaded has to include the following: title • Marks WM, Filly RA, Callen PW. Real-time evaluation of
of the manuscript, abstract and keywords, main manuscript, pleural lesions: new observations regarding the probability
references, tables, legend. Do not insert here the name of the of obtaining free fluid. Radiology 1982;142:163-164.
authors and affiliation in order to ensure a blind review. b) Papers published only with DOI numbers:
b) in step 4 of submission process the supplementary files • Guerriero S, Alcazar JL, Pascual MA, Ajossa S, Olartecoe-
that must be uploaded are: title page, images, submission let- chea B, Hereter L. The pre-operative diagnosis of metastatic
ter, declaration of conflict of interest. ovarian tumors is related to the origin of the primary tumor.
Ultrasound Obstet Gynecol 2011, doi: 10.1002/uog.10120.
Title page includes: title of the paper, full names of the c) Book:
authors, department and institution(s) where the study was • Talano JV, Gardin JM. Textbook of two dimensional echo-
conducted, postal code, city, district, phone and/or fax number cardiography. London: Gruene & Stratton, 1983.
and/or e-mail address for contacting the first author and cor- d) Book chapter:
responding author, full postal address for correspondence and • Brooks M. The Liver. In: Goldberg BB, Pettersson H (eds).
ordering reprints. Ultrasonography. Oslo, The Nicer Year Book 1996:55-82.
Abstract (on a separate page) preceding the text body. Tables should be added at the end of the main document,
In the case of original papers, abstracts should not exceed in the same word document.
250 words and should have the following structure: 1) aims;
2) material and methods; 3) results; 4) conclusions. Abstracts 4. Editorial policy
for literature reviews and educational papers should not
exceed 200 words. For case reports, the abstract must not Medical Ultrasonography promotes evaluation of all the
exceed 120 words and must underline the following: 1) pur- scientific papers by independent reviewers. The editor-in-
pose of the presentation; 2) peculiarities of the case; ranking chief or one of the editors evaluates each manuscript and, in
of the issues approached within the general knowledge of the 1-3 weeks from reception, decides upon their priority level
respective condition. (sent to review, rejected without being sent for review or re-
Three to five keywords must be selected for every paper turned to authors with suggestions for improvement before
from the Index Medicus (http://www.ncbi.nlm.nih.gov/sites/ submitting to review). The editors reserve the right to request
entrez?db=mesh); the key words should be inserted after the any changes they may consider appropriate, in the title, struc-
abstract and separated by semi-colon (term1; term2; term3). ture or body of the paper.
The papers are submitted to two blind reviewers with
The main document has to be structured as follows: expertise in ultrasonography. Based on the reviewers’ recom-
Introduction – should define the topic of the paper and mendations, the editors decide whether a paper is published
present the status of current knowledge in the field. or not. In case of marked discrepancy between the decisions
Material and methods – should describe the equipment of the two reviewers, the editor may send the manuscript to
employed, the group of patients studied and the methodology. another arbitrator for additional comments and a recommend-
We recommend specification of the type of ultrasound ed decision. The full decisional process may last 6-8 weeks.
equipment employed. The statistical analysis methodology Failure of the authors to comply with the editorial revision
used must also be described. requests may induce publication rejection.
Results – should present the obtained data, in a concise A 150 euro processing fee (100 euro for the first author if
manner, preferably in tables and diagrams. SRUMB member) is charged for each accepted paper starting
Discussions – should present the interpretation of the re- from 1st March 2021. The fee must be paid ahead of publication,
searcher’s results from the perspective of relevant literature at the time when the authors give the authorization for publi-
data. cation of the final proof. Instructions for online payment are
Conclusions of the paper must be clearly stated in the end. available on the journal site. No free reprints of the published
Acknowledgements – should be made only to those who paper are supplied. No additional reprints may be ordered.
have made a substantial contribution to the study
References must include only papers that are quoted in the 5. Submitting manuscripts for publication
text and that have been published. References must be num-
bered in Arabic numerals in the order in which they appear To submit a manuscript for publication, one author
in the text (where they should be inserted between square should register on our website, and follow the five steps of the
brackets [ ]) and listed in numerical order. Titles of medical submission process. The journal does not accept email or
journals must be abbreviated according to the Index Medicus. CD submissions of articles. The editorial office will send a
All authors must be quoted for an article, if they are up to confirmation e-mail to the correspondence address.
six. Over seven authors, only the first three will be quoted, Remark: Medical Ultrasonography cannot be held respon-
followed by the “et al” indication. References should be listed sible for losing or damaging the files delivered through the
according to the following format (examples): Internet.