• Nu S-Au Găsit Rezultate

Added value of two-dimensional shear wave elastography to ultrasonography for staging common femoral vein thrombi

N/A
N/A
Protected

Academic year: 2022

Share "Added value of two-dimensional shear wave elastography to ultrasonography for staging common femoral vein thrombi"

Copied!
8
0
0

Text complet

(1)

Original papers

DOI: 10.11152/mu-926

Added value of two-dimensional shear wave elastography to ultrasonography for staging common femoral vein thrombi

Fu-shun Pan, Wen-shuo Tian, Jia Luo, Ming Liu, Jin-yu Liang, Ming Xu,Yan-ling Zheng, Xiao-yan Xie

Department of Medical Ultrasonics, Institute for Diagnostic and Interventional Ultrasound, the First Affiliated Hospi- tal, Sun Yat-Sen University, Guangzhou, P.R. China

Received 01.07.2016 Accepted 25.09.2016 Med Ultrason

2017, Vol. 19, No 1, 51-58

Corresponding author: Xiao-yan Xie

Department of Medical Ultrasonics, Institute for Diagnostic and interventional Ultrasound, the First Affiliated Hospital,

Sun Yat-Sen University

58 Zhongshan Road 2, Guangzhou 510080, P.R. China

Phone and Fax: 86-20-87765183 E-mail: [email protected]

Introduction

Deep venous thrombosis (DVT) is a common disor- der with potential complications including pulmonary embolism and postthrombotic syndrome [1]. Compres- sion ultrasonography (US) is widely used to establish a DVT diagnosis, with a sensitivity and specificity of 95%

and 98%, respectively, for the detection of proximal leg veins thrombi [2]. However, another important issue is

the therapeutic strategy, which highly depends on the thrombi age, acute thrombi being initially treated with anticoagulant, whereas chronic thrombi may need no therapy [3,4]. It is a widespread approach to use the onset of symptoms for defining DVT age [3-6]. Surprisingly, most patients who develop a DVT are asymptomatic [7].

Thus, it is important for the clinician to know the DVT age. Although US parameters (vein diameter and echo- genicity of thrombi) are helpful for staging DVT, the optimal diagnostic threshold is still controversial due to wide disparity in different studies [6,8,9]. Additionally, it is also challenging to stage thrombi with other imaging modalities such as magnetic resonance (MR), computed tomography (CT), and nuclear techniques [10-14].

Recently, two-dimensional shear wave elastogra- phy (2D-SWE) has emerged as a novel elastography technique, that allows quantitative analysis of stiffness expressed as Young’s modulus (kPa) with high reliabil- ity and reproducibility [15]. It is well documented that Abstract

Aims: To evaluate whether adding two-dimensional shear wave elastography (2D-SWE) to ultrasonography (US) could improve the performance for staging common femoral vein thrombi (CFVT). Material and methods: A total of 194 consecu- tive patients with CFVT who underwent US and 2D-SWE were enrolled. These patients were categorized into three groups according to CFVT duration: Stage A (≤14 days), Stage B (14 days to 6 months), and Stage C (≥6 months). The diagnostic performance was assessed by the area under the receiver operating characteristic curve (AUC). Results: Among all US fea- tures, CFV diameter ratio of thrombosed leg to contralateral leg (CFVD_ratio) showed the highest AUC in predicting Stage A and Stage C (0.87 and 0.84, respectively). The diagnostic performance of 2D-SWE value of CFVT (CFVT_E) is comparable with that of CFVD_ratio for Stage A (AUC: 0.85, p=0.630), whereas inferior to that of CFVD_ratio for Stage C (AUC: 0.73, p=0.026). Combining CFVD_ratio with CFVT_E showed lower performance in predicting Stage A (AUC: 0.81, p=0.021) and Stage C (AUC: 0.67, p<0.0001) relative to CFVD_ratio alone. However, this combination increased the specificity from 80.3% to 92.7% (p<0.0001) without a significant reduction of sensitivity (from 77.2% to 70.2%, p=0.371) for predicting Stage A. Conclusions: Adding 2D-SWE to US did not improve the diagnostic performance for staging CFVT compared with US alone. However, the combination improved the specificity in predicting CFVT less than 14 days without loss of sensitivity.

Keywords: elastography, ultrasonography, femoral vein, thrombosis

(2)

thrombi harden as they age [16-18], the hardening pro- cess is largely due to multiple factors, including organi- zation of the thrombi and the clot transformation from one composed mainly of platelets to one composed main- ly of fibrin. However, these studies were based on either animal or experimental models and had analyzed the evolution of thrombi elasticity only in acute stages. Few studies have evaluated US elasticity imaging of thrombi in humans [6,19]. Moreover, all the thrombi elasticity re- sults of these studies were not based on 2D-SWE [6,16- 19]. A recent study used 2D-SWE to evaluate the elastic- ity of venous thrombi [20], however, the results were also based on animal model.

To date, no study has reported the elasticity features of 2D-SWE in different stages of the common femoral vein thrombosis (CFVT) in humans. The purpose of this study was to evaluate whether adding 2D-SWE to US could improve the performance for staging CFVT.

Materials and methods Patients

This prospective study was approved by the Insti- tutional Review Board, and informed consent was ob- tained from all the subjects. Between January 2014 and July 2015, 289 consecutive patients (131 male, 158 female; mean age of 48.9±13.2 years, range of 17-94 years) with CFVT diagnosed by compression US were enrolled. The indications for vein US were as follows:

a) swelling of lower limb; b) risk factors for DVT such as trauma, cancer, prolonged immobility, and oral con- traceptive therapy; c) CFVT detected via US at a sec- ondary or junior clinic; and d) follow-up examination of an existing CFVT, previously diagnosed by US. Of these, 95 patients were excluded for the following rea- sons: a) bilateral CFVT (n=11 ); b) CFVT thickness less than 2.0 mm (n=25 ) because the minimal region of inter- est (ROI) of 2D-SWE was of 2.0 mm; c) inconclusive evidence of CFVT duration due to either uncertain onset date of symptoms (n=13) or asymptomatic presentation (n=10); d) suspicion of recurrent CFVT (n=7) such as new symptoms of acute DVT or enlargement of thrombi thickness (≥2 mm) compared with those available from the previous US examination [21]; e) prior thrombolytic treatment (n=13); and f) failure of either US (n=0) or 2D- SWE (n=16) acquisition. Ultimately, 194 patients (105 women and 89 men; 48.8±16.9 years; range 19-94 years) were enrolled in the final analysis. The CFVT age was defined as the duration from the onset of symptoms to presentation, either directly from examination or from the previous CFVT US examinations. These patients en- rolled were categorized into three groups according to

CFVT age: Stage A (≤14 days, 57 patients), Stage B (14 days to 6 months, 60 patients), and Stage C (≥6 months, 77 patients). The average CFVT age was of 5.6±2.8 days (range 1-11 days) for Stage A, 67.1±41.0 days (range 16- 165 days) for Stage B, and of 57.4±71.4 months (range 6.5-288 months) for Stage C.

Conventional US and 2D-SWE imaging

All US and 2D-SWE were performed using the same US machine (SuperSonic Imagine, Aix-en-Provence, France) with a 15-4 MHz linear probe by the same op- erator (F.S.P., with 5 years of experience in vascular US and 3 years of experience in US elastography) who was blinded to participant history. For each patient, all the US and 2D-SWE were performed on the same day. The US and 2D-SWE images of the targeted CFVT were ob- tained near the greater saphenous vein junction if throm- bi were present at this level; otherwise, acquisitions were obtained at the most proximal level in the CFV contain- ing clot.

To confirm the diagnosis of DVT and the extent of thrombi, the compression US was performed from the level of the common iliac vein to the calf veins in both transverse and longitudinal planes, including the femoral vein, the popliteal vein, and the great saphenous veins.

Color and spectral Doppler US were applied to confirm whether obvious blood flow surrounding thrombi was detectable. All the grayscale US settings were standard- ized as follows: dynamic range 50, gain 40%, time gain compensation curve positioned at the center, and focus position at the bottom of CFV.

The 2D-SWE of CFVT was subsequently performed in the same longitudinal plane as US. The 2D-SWE set- tings were standardized as follows: gain 70%, option penetration, elasticity range 100 kPa, and smallest size of the color box, both 1.1 cm in width and length. When 2D-SWE mode was activated, the transducer was held steady with no movement or compression and the 2D- SWE box was moved to target the CFVT. The image was frozen and saved after a few seconds of immobilization.

2D-SWE acquisition was considered to be a failure when the signal was obtained in less than half of the 2D-SWE box. Three valid and consecutive 2D-SWE images were acquired in the same imaging plane for each patient.

Image interpretation

All the images were digitally stored on the US sys- tem for subsequent offline analysis. One blinded radi- ologist (W.S.T.) with 2 years of experience in 2D-SWE reviewed the US and 2D-SWE features. Identification of the patients and clinical results were not available to the investigator.

The targeted CFVT was evaluated for occlusion de- gree (totally or partially occluded), depth (vertical dis-

(3)

tance from the skin to the anterior wall of CFV), and CFVD_ratio (CFV diameter ratio of thrombosed leg to contralateral leg). Measurements of thrombosed CFV di- ameter were made by the use of transverse plane of CFV near the greater saphenous vein junction if thrombi were present at this level; otherwise, measurements were ob- tained at the most proximal level in the CFV containing clot. Measurements of the contralateral CFV diameter were obtained at the level similar to thrombosed CFV.

Additionally, CFVT_echo (echo of CFVT) and Echo_ra- tio (echo ratio of CFVT to fat tissue) were quantitatively evaluated using Image J software (version 1.44p; Na- tional Institutes of Health, Bethesda, MD). A ROI with the size and shape fitted to the thrombi was manually drawn and the quantitative echogenicity of thrombi was obtained. Another ROI, with a circular shape of a diam- eter of 5 mm, was positioned at the layer of the fat tissue to obtain the corresponding echo value.

The size and the location of the ROI for 2D-SWE was standardized as follows: 1) the vessel wall and the obvious blood flow surrounding thrombi were avoided;

2) a proximal floating thrombus tip located at CFV was avoided; and 3) the number and site of ROIs for each im- age varied depending on the size of the thrombi. If two or more ROIs were applied, they should be placed without overlap. Thus, the number of ROIs for each image ranged from one to five because the width of 2D-SWE color box was of 1.1cm (fig 1-3). The average spatial mean-value of all the ROIs measurements was used as the representa- tive value of the CFVT for each acquisition. The average value of the three consecutive acquisitions was calculat- ed as the ultimate elastic modulus of CFVT (CFVT_E).

Statistical analysis

Statistical analysis was performed using MedCalc Software, version 11.2 (MedCalc program, Belgium).

The one-sample Kolmogorov-Smirnov test was used to test the normal distribution of the quantitative variables.

When the quantitative variables were normally distrib- uted, the results were described as mean values and standard deviations; otherwise, medians and interquartile ranges (IQR; 25th-75th percentile) were reported. The

qualitative variables were summarized as counts and per- centages. Comparisons of clinical as well as US charac- teristics of the three groups were performed using either X2 test or Fisher’s exact test for categorical variables and ordinary two-way ANOVA analysis of variance by ranks for continuous variables. Spearman’s coefficient was used to test the correlation between the CFVT stage and assay results. The diagnostic performance of US and 2D- SWE, as well as their combinations was assessed using receiver operating characteristic (ROC) curves and the area under the ROC curves (AUCs) using the following parameters: Stage A versus Stage B-C (=Stage A), and Stage A-B versus Stage C (=Stage C). The AUCs were compared by using the method proposed by DeLong et al [22].The ROC curve analyses were also used to deter- mine the optimal diagnostic cut-off value for US and 2D- SWE parameters, and the corresponding sensitivity and specificity were calculated, which were compared using the McNemar test. Both point estimates and 95% confi- dence intervals (CIs) were used. Statistical significance was defined as p<0.05 for two-tailed tests.

Results

Basic characteristics

The success rate of 2D-SWE was 92.4% (194 of 210). The reasons for 2D-SWE failure were as follows:

acoustic shadowing due to scarring from either CFV drug injection (n=1) or previous inguinal surgical inter- vention (n=2), large CFVT depth due to obesity (n=10, body mass index (BMI) >32 kg/m2 for all), inguinal he- matoma (n=2) or femoral artery pseudoaneurysm (n=1).

Among the 194 patients, thrombi were detected in the iliac vein of 122 patients, in the superficial femoral vein of 155 patients, in the popliteal vein of 104 patients and in the calf vein of 113 patients. There was no difference between the groups in regard to age, gender distribution, BMI, the percentage of left side CFVT and CFVT depth, except the percentage of CFV completely occluded by thrombi and patients receiving anticoagulation treatment (Table I).

Table I. Basic characteristics of study patients

Characteristics Stage A (n=57) Stage B (n=60) Stage C (n=77) p value

Age (y) 49.8±18.8 47.2±15.2 49.3±16.9 0.630

Men: n (%)§ 26(45.6) 31(51.7) 32(41.6) 0.499

BMI (kg/m2) 22.8±2.7 22.7±2.8 22.6±3.0 0.826

Left CFVT: n (%)§ 42(73.7) 48(80.0) 59(76.6) 0.608

CFVT depth (cm)

CFV totally occluded: n(%)§ 17.8±4.1

43(75.4) 17.5±4.8

12(20.0) 18.4±5.1

5(6.5) 0.545

<0.001

Anticoagulation treatment: n(%)§ 15(26.3) 49(81.7) 72(93.5) <0.001

Unless otherwise indicated, results are expressed as mean ± standard deviation. § Data are numbers of patients, with percentages in paren- theses.

(4)

Conventional US and 2D-SWE features

The US and 2D-SWE characteristics of the study patients are presented in Table II. The CFVD ratio was negatively correlated with CFVT stage (r=-0.659, p<0.0001). There was only a weak positive correla- tion between CFVT_echo, Echo_ratio and CFVT stage (r=0.295, p<0.0001; and r=0.339, p<0.0001, respective- ly) The medians of CFVT_E were 8.2 kPa for Stage A,

17.1 kPa for Stage B and 21.5 kPa for Stage C. A positive correlation was observed between CFVT_E and CFVT stage (r=0.536, p<0.0001) (fig 4).

Diagnostic performance of conventional US and quantitative 2D-SWE features

The diagnostic performance of US and 2D-SWE features in predicting CFVT stages are detailed in Table III and figure 5. For predicting Stage A, the Fig 1. Longitudinal imaging of a left CFVT in a 66-year-old

man with lower limb swelling for 3 days prior to presentation.

Top: 2D-SWE image shows a homogeneous blue pattern with an elasticity of 6.3 kPa. Bottom: B-mode image shows the CFV contains hypoechoic thrombi.

Fig 2. Longitudinal imaging of a right CFVT (proven 2 months ago by compression US) in a 53-year-old woman. Top: 2D- SWE image shows a relatively homogeneous blue and purple pattern with an ultimate elasticity (average mean-value of the two measurements) of 11.9 kPa. Bottom: B-mode image shows that the CFV contains heterogeneous thrombi with partial re- canalization.

Fig 3. Longitudinal imaging of a left CFVT (proven 9 months ago by compression US) in a 40-year-old man. Top: 2D-SWE image shows a heterogeneous red and yellow pattern with an ultimate elasticity (average mean-value of the three measure- ments) of 51.7 kPa. Bottom: B-mode image shows that the CFV contains heterogeneous thrombi with partial recanaliza- tion.

Fig 4. Box-and-whisker plots of CFVD_ratio, CFVT_echo, Echo_ratio, and CFVT_E in study population according to CFVT stage. The top and bottom of each box are 75th and 25th percentiles, respectively; the horizontal line in each box is the median (50th percentile); and the top and bottom of the whisk- ers are minimum and maximum, respectively.

(5)

AUCs of CFVD_ratio were higher compared to CFVT_echo and Echo_ratio (p<0.01 for both), but no significant difference was observed compared to CFVT_E (p=0.630). For predicting Stage C, the AUCs of CFVD_ratio were the highest compared to CFVT_echo, Echo_ratio, and CFVT_E (p< 0.01 for all). However, contrary cutoff values were obtained for CFVT_echo and Echo_ratio.

Diagnostic performances of combined conventional US and 2D-SWE features

Due to contrary cutoff values for CFVT_echo and Echo_ratio in predicting Stage A and Stage C, both pa- rameters cannot be regarded as a useful indicator for staging CFVT; thus we combined CFVD_ratio and CFVT_E to assess the diagnostic performance for stag- ing CFVT (Table IV). The AUCs were of 0.81 in pre- dicting Stage A and of 0.67 in predicting Stage C, these values showed lower performance compared to CFVD_

ratio alone (p=0.0208 for Stage A, and p<0.0001 for Stage C). Meanwhile, this combination did not increase the specificity significantly (from 77.8% to 85.5%,

p=0.216) and decreased the sensitivity (from 77.9% to 49.4%, p< 0.0001) for predicting Stage C. However, this combination increased the specificity from 80.3%

to 92.7% (p<0.0001) without a significant reduction of sensitivity (from 77.2% to 70.2%, p=0.371) for predict- ing Stage A.

Fig 5. ROC curves of CFVD_ratio, CFVT_E alone or com- bined in predicting (a) Stage A (Stage A vs. Stage B-C) and (b) Stage C (Stage A-B vs. Stage C).

Table II. Conventional US and 2D-SWE characteristics of study patients

Characteristics Stage A (n=57) Stage B (n=60) Stage C (n=77) Correlation coefficient CFVD_ratio 1.44±0.39

(1.34-1.55) 1.07±0.24

(1.00-1.13) 0.85±0.21

(0.80-0.90) -0.659

p< 0.0001

CFVT_echo 37.2±17.1

(31.1-40.2) 43.4±19.5

(33.9-50.0) 52.2±22.9 (40.6-54.9) 0.295 p< 0.0001 Echo_ratio 0.41±0.18

(0.36-0.46) 0.45±0.17

(0.40-0.50) 0.57±0.21 (0.50-0.60) 0.339 p< 0.0001 CFVT_E (kPa)§ 8.2 (5.8-11.2)

[7.1-8.8] 17.1 (13.4-22.6)

[14.6- 20.2] 21.5 (13.7-32.8)

[18.2-25.2] 0.536

p< 0.0001

Except for the correlation coefficient and unless otherwise indicated, data are means ± standard deviation and data in parentheses are 95%

CIs. § Data are medians, data in parentheses are interquartile range (25th–75th percentile), and data in brackets are 95% CIs. kPa- kilopascal.

Table III. Optimal cut-off values and areas under the receiver operating characteristic curve of US and 2D-SWE for predicting Stage A and Stage C

Parameters cutoff AUCs (95%CI) p* value for comparison of AUCs

Stage A

(1) CFVD_ratio > 1.16 0.87 (0.81-0.91)

(2) CFVT_echo ≤ 48.2 0.65 (0.58-0.72) < 0.001

(3) Echo_ratio ≤ 0.4 0.65 (0.58-0.71) < 0.001 0.804

(4) CFVT_E (kPa)

(1) + (4) ≤ 13.9 0.85 (0.79-0.90)

0.81 (0.75-0.87) 0.630

0.021 < 0.001

0.002 < 0.001

0.003 0.228

Stage C

(1) CFVD_ratio ≤ 0.98 0.84 (0.78-0.89)

(2) CFVT_echo > 37.7 0.66 (0.59-0.72) < 0.001

(3) Echo_ratio > 0.3 0.69 (0.62-0.76) 0.002 0.445

(4) CFVT_E (kPa)

(1) + (4) > 17.2 0.73 (0.66-0.80)

0.67 (0.60-0.74) 0.026

< 0.001 0.146

0.732 0.392

0.735 0.072

CI – confidential interval. *Only AUCs were compared. The first column P value was compared to CFVD_ratio; the second column P value was compared to CFVT_echo; the third column P value was compared to Echo_ratio; the fourth column P value was compared to CFVT_E.

(6)

Discussions

According to the American College of Chest Physi- cian (ACCP) guidelines, for patients diagnosed with acute iliofemoral DVTs with symptoms of less than 14 days, thrombolysis is specifically recommended instead of simple anticoagulation; duration of anticoagulant ther- apy may be of 3-6 months, or indefinite, depending on the risk factors of DVT [3]. In the present study we chose a 14 day period to define Stage A, also named as acute thrombi, and a cut-off of 6 month period to differentiate between Stages B and C. However, most patients who develop DVT are asymptomatic [7]. With the broad ap- plication of US, we believe that the frequency of detect- ing asymptomatic CFVT may be increasing. Thus, it is important for the clinician to know the CFVT age, espe- cially whether it exceeds 14 days.

Fowlkes et al [9] reported that thrombi echogenicity is a useful feature for staging thrombi. However, their study was based on an experimental model, and the anal- ysis time point was less than 14 days after thrombi initia- tion. Rubin et al [6] assessed the relative thrombi echo- genicity of acute (<14 days) and chronic (>8 months) DVTs and revealed echogenicity difference between the two groups was highly significant, with AUC of 0.92.

Conversely, Murphy et al [8] performed serial US fol- low-ups up to 6 months after the initial DVT diagnosis and semi-quantitatively analyzed the echogenicity in dif- ferent patients. Their findings suggested that echogenic- ity is not reliable for staging thrombi. Several reasons might explain the inconsistent results of different studies.

One reason might be the different enrollment criteria of thrombi age; another reason might be the system inde-

pendent measurement of echogenicity which highly de- pends on the transducer frequency, thrombi depth, gain level of grayscale US, etc. In the present study, patients with CFVT were systematically enrolled in different stages, which ranged from 1 day to 288 months. Addi- tionally, the grayscale US settings were standardized for all patients to eliminate measurement biases. We quanti- tatively analyzed CFVT_echo and Echo_ratio at different CFVT stages. Interestingly, contrary cutoff values were obtained for CFVT_echo (≤48.2 for Stage A, and >37.7 for Stage C) and Echo_ratio (≤0.4 for Stage A, and >0.3 for Stage C) in predicting Stage A and Stage C.

Murphy et al [8] also mentioned that venous diam- eter was useful in discriminating acute from chronic DVTs. However, they only evaluated diameter of throm- bosed vein, contralateral vein diameter being ignored.

That is, individual variations were not evaluated. Fur- thermore, the study did not obtain information regarding diagnostic performance. In the present study, we evalu- ated the CFVD ratio to eliminate individual variation and analyzed the performance in predicting CFVT stage.

The AUCs were 0.87 for Stage A and 0.84 for Stage C, indicated that CFVD ratio was useful for distinguishing CFVT stage.

To date, this is the first study assessing CFVT elastic- ity in humans using 2D-SWE. Several studies [16-18,20]

based on experimental models reported that the elasticity of acute thrombi had increased gradually over time. An- other study [19] evaluated only two patients in different stages of CFVT. Rubin et al [6] also evaluated the throm- bi elasticity and showed a high AUC of 0.97 to discrimi- nate acute (<14 days) from chronic (>8 months) DVTs.

Moreover, all the thrombi elasticity results of these stud- ies were not based on 2D-SWE technique, except the study based on experimental models by Mfoumou et al [20]. In the present study, the AUCs were of 0.85 in pre- dicting Stage A and of 0.76 in predicting Stage C, which were significantly lower than the previous study reported [6]. Patients with thrombi aged between 14 days and 8 months were outside the scope of their study, which might be an important explanation for the inconsistent results. Additionally, different measurement system of elasticity might be another reason.

Many studies have focused on other imaging modali- ties in attempting to stage thrombi. MR imaging may be helpful in distinguishing acute from chronic DVTs [10- 12]; however, these findings were either preliminary or non-specific. With the emergence of nuclear medicine, Brighton et al [13] explored that the uptake image of

99mTc-rt-PA into the thrombi might have a potential role for staging thrombi. However, none of these nuclear techniques have been widely adopted into clinical prac- Table IV. Optimal cut-off values of conventional US and

SWE for the diagnosis of CFVT age according to the highest Youden’s index (sensitivity + specificity -1)

Parameters Se (%) Sp

(%) PPV (%) NPV

(%) Ac (%) Stage A

(1) CFVD_ratio 77.2 80.3 62.0 89.4 79.4 (2)CFVT_echo 78.9 46.0 37.8 84.0 57.1 (3)Echo_ratio 64.9 56.2 38.1 79.4 58.8

(4)CFVT_E 82.5 73.0 56.0 90.9 75.8

(1) + (4) 70.2 92.7 80.0 88.2 90.2

Stage C

(1) CFVD_ratio 77.9 77.8 69.8 84.3 77.8 (2)CFVT_echo 68.8 55.6 50.5 73.0 60.8 (3)Echo_ratio 86.7.9 34.2 47.6 85.1 57.2

(4)CFVT_E 67.5 70.9 60.5 76.9 68.4

(1) + (4) 49.4 85.5 70.2 65.8 78.9

Se – Sensitivity, Sp – Specificity, PPV – positive predictive value, NPV – negative predictive value, Ac – Accuracy

(7)

tice. Furthermore, it is difficult to differentiate acute from chronic DVTs via CT scans or venography [14,23]. Thus, it is challenging to stage thrombi for various imaging modalities available.

In view of this situation, a test with high specific- ity for staging thrombi is more important for a clinician who has to make an urgent treatment decision. Because US and 2D-SWE can be implemented using the same machine, combining them could be regarded as a one- stop examination, which would make the simultaneous diagnosis and staging of the thrombi possible. US and 2D-SWE should be the preferred modalities due to their low cost, performance, elimination of radiation, and availability on bed-side. However, early stage DVT has potential complications, including pulmonary embolism and postthrombotic syndrome, thus a decrease of sensi- tivity of a diagnostic method is also not acceptable. In this study, we attempted to combine CFVD ratio with 2D-SWE for staging CFVT. If both methods were con- cordant, the AUCs were of 0.81 in predicting Stage A and of 0.67 in predicting Stage C, which both showed lower performance compared to CFVD ratio alone.

Meanwhile, this combination did not significantly in- crease the specificity and decreased the sensitivity for predicting Stage C. However, we obtained a higher specificity and no significant reduction of sensitivity for predicting Stage A. In other words, combining 2D-SWE with US did not improve the diagnostic performance for staging CFVT compared to US alone. However, it is highly specific for predicting CFVT stage A without loss of sensitivity.

This study had several limitations. First, placing of ROIs was dependent on B-mode ultrasound criteria and thus might had suffered in cases of low echo signal dif- ferences such as between hypoechoic thrombi and blood.

This might have led to falsely low values of CFVT_E.

Second, up to 70.1% (136/194) of this population re- ceived anticoagulation treatment; it was conceivable that untreated thrombi might have not behaved similarly to a treated thrombi. Third, owing to a relatively high probe used in our study, it limited the resolution on B-mode US as well as 2D-SWE. Thus, we only assessed patients with CFVT. Although studies have revealed that iliofemoral DVT was more common than calf DVT, and no patholog- ical or structural differences were associated with throm- bi location [5], no generalizations could be made until DVTs involving other locations were studied. Fourth, this study was executed in our single institution; thus, our findings need to be validated in larger multicenter trial.

In conclusion, adding 2D-SWE to US did not improve the diagnostic performance for staging CFVT compared with US alone. However, it is highly specific for predict-

ing Stage A (less than 14 days) without loss of sensitivity, suggesting that 2D-SWE may be a promising adjunctive tool for risk stratification and therapy planning.

Conflict of interest: none

Reference

1. Peterson KL. Acute pulmonary thromboembolism: has its evolution been redefined? Circulation 1999;99:1280-1283.

2. Lensing AW, Prandoni P, Brandjes D, et al. Detection of deep-vein thrombosis by real-time B-mode ultrasonogra- phy. N Engl J Med 1989;320:342-345.

3. Kearon C, Kahn SR, Agnelli G, et al. Antithrombotic ther- apy for venous thromboembolic disease: American Col- lege of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133:454S-545S.

4. Wakefield TW. Treatment options for venous thrombosis. J Vasc Surg 2000;31:613-620.

5. Kyrle PA, Eichinger S. Deep vein thrombosis. Lancet 2005;365:1163-1174.

6. Rubin JM, Xie H, Kim K, et al. Sonographic elasticity im- aging of acute and chronic deep venous thrombosis in hu- mans. J Ultrasound Med 2006;25:1179-1186.

7. Fraser JD, Anderson DR. Deep venous thrombosis: recent advances and optimal investigation with US. Radiology 1999;211:9-24.

8. Murphy TP, Cronan JJ. Evolution of deep venous throm- bosis: a prospective evaluation with US. Radiology 1990;177:543-548.

9. Fowlkes JB, Strieter RM, Downing LJ, et al. Ultrasound echogenicity in experimental venous thrombosis. Ultra- sound Med Biol 1998;24:1175-1182.

10. Froehlich JB, Prince MR, Greenfield LJ, Downing LJ, Shah NL, Wakefield TW. “Bull’s-eye” sign on gadolini- um-enhanced magnetic resonance venography determines thrombus presence and age: a preliminary study. J Vasc Surg 1997;26:809-816.

11. Spritzer CE, Trotter P, Sostman HD. Deep venous thrombo- sis: gradient-recalled-echo MR imaging changes over time- -experience in 10 patients. Radiology 1998;208:631-639.

12. Erdman WA, Jayson HT, Redman HC, Miller GL, Par- key RW, Peshock RW. Deep venous thrombosis of ex- tremities: role of MR imaging in the diagnosis. Radiology 1990;174:425-431.

13. Brighton T, Janssen J, Butler SP. Aging of acute deep vein thrombosis measured by radiolabeled 99mTc-rt-PA. J Nucl Med 2007;48:873-878.

14. Nazaroglu H, Ozmen CA, Akay HO, Kilinc I, Bilici A. 64- MDCT pulmonary angiography and CT venography in the diagnosis of thromboembolic disease. AJR Am J Roentgen- ol 2009;192:654-661.

15. Bercoff J, Tanter M, Fink M. Supersonic shear imaging:

a new technique for soft tissue elasticity mapping. IEEE Trans Ultrason Ferroelectr Freq Control 2004;51:396-409.

(8)

16. Xie H, Kim K, Aglyamov SR, et al. Staging deep venous thrombosis using ultrasound elasticity imaging: animal model. Ultrasound Med Biol 2004;30:1385-1396.

17. Emelianov SY, Chen X, O’Donnell M, et al. Triplex ultra- sound: elasticity imaging to age deep venous thrombosis.

Ultrasound Med Biol 2002;28:757-767.

18. Geier B, Barbera L, Muth-Werthmann D, et al. Ultrasound elastography for the age determination of venous throm- bi. Evaluation in an animal model of venous thrombosis.

Thromb Haemost 2005;93:368-374.

19. Rubin JM, Aglyamov SR, Wakefield TW, O’Donnell M, Emelianov SY. Clinical application of sonographic elastic- ity imaging for aging of deep venous thrombosis: prelimi- nary findings. J Ultrasound Med 2003;22:443-448.

20. Mfoumou E, Tripette J, Blostein M, Cloutier G. Time-de- pendent hardening of blood clots quantitatively measured in vivo with shear-wave ultrasound imaging in a rabbit model of venous thrombosis. Thromb Res 2014;133:265-271.

21. Prandoni P, Cogo A, Bernardi E, et al. A simple ultrasound approach for detection of recurrent proximal-vein thrombo- sis. Circulation 1993;88:1730-1735.

22. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biomet- rics 1988;44:837-845.

23. Hull RD, Carter CJ, Jay RM, et al. The diagnosis of acute, recurrent, deep-vein thrombosis: a diagnostic challenge.

Circulation 1983;67:901-906.

Referințe

DOCUMENTE SIMILARE

Aim: To evaluate the range of liver stiffness (LS) cut-off values for predicting different stages of liver fibrosis (LF) for 2D-SWE-GE implemented on three different systems

Shear wave elastography (SWE) is the most cut- ting-edge ultrasonic elastography technology at present. Compared with strain elastography, SWE has the advan- tages of

Aim: To evaluate the performance of 2D shear-wave elastography from General Electric (2D-SWE.GE) for the nonin- vasive assessment of liver fibrosis and to identify liver stiffness

2D SWE was performed with either the SC6-1 or SL10-2 probe through an intercostal space with patients in the supine position and regular free breathing. We used the same

According to our previous investigations it seems that tolerance, whether regarded as a political practice or a philosophical or moral principle, is a strategy (or tactics) of one

Aim: To evaluate the ultrasound (US) modifications [grey scale, Doppler, 2D-share wave elastography (2D-SWE)] of salivary (parotid and submandibular) and lacrimal glands in

The purpose of this study was thus to compare the Intra-/Inter-observer variation and accuracy in vitro of point shear wave elastography (pSWE) with that of 2D-SWE and to

Aims: To evaluate the differences and changes of testicular volume and elasticity in normal (NL) testes and undescended testes (UDTs) of children using shear wave elastography