- Email: [email protected]
Value of Shear Wave Elastography in the Diagnosis of Gouty and Non-Gouty Arthritis
Ultrasound in Med. & Biol., Vol. -, No. -, pp. 1–9, 2017 Ó 2016 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/$ - see front matter
http://dx.doi.org/10.1016/j.ultrasmedbio.2016.12.012
d
Original Contribution VALUE OF SHEAR WAVE ELASTOGRAPHY IN THE DIAGNOSIS OF GOUTY AND NON-GOUTY ARTHRITIS YUANJIAO TANG,* FENG YAN,*y YUJIA YANG,* XI XIANG,* LIYUN WANG,* LINGYAN ZHANG,* and LI QIU* * Department of Ultrasound, West China Hospital of Sichuan University, Chengdu, Sichuan, China; and y Clinical Ultrasound Imaging Drug Research Lab, West China Hospital of Sichuan University, Chengdu, Sichuan, China (Received 22 August 2016; revised 15 December 2016; in final form 20 December 2016)
Abstract—Our aim was to analyze the diagnostic performance of shear wave elastography (SWE) in the diagnosis of gouty arthritis (GA) and non-gouty arthritis (non-GA). Thirty-nine patients in the GA group and 55 patients in the non-GA group were included in the study. Based on the echo intensity of the joint lesions, the GA group was subdivided into hypo-echoic GA, slightly hyper-echoic GA and hyper-echoic GA subgroups. Quantitative SWE features were evaluated and receiver operating characteristic analysis was performed. On the basis of the study, the elastic modulus (Emax), mean elastic modulus (Emean), minimum elastic modulus (Emin) and elastic modulus standard deviation (ESD) were significantly higher in the GA group than in the non-GA group and were highest in the hyper-echoic GA subgroup (p , 0.01 for all). Emin, Emean and Emax were significantly higher in the hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD was significantly higher in the hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all). Based on the hypo-echoic GA subgroup and non-GA group, areas under the receiver operating characteristic curves for the prediction of GA were 0.749 for Emin, 0.877 for Emean, 0.896 for Emax and 0.886 for ESD, with optimal cutoff values of 29.40 kPa for Emin, 45.35 kPa for Emean, 67.54 kPa for Emax and 7.85 kPa for ESD. Our results indicate that SWE can differentially diagnose GA and non-GA, especially when the ultrasound manifestations are not typical. (E-mail: [email protected]) Ó 2016 World Federation for Ultrasound in Medicine & Biology. Key Words: Gouty arthritis, Non-gouty arthritis, Ultrasound, Shear wave elastography, Stiffness.
increased with improved living standards (Chilappa et al. 2010). The most reliable diagnostic criterion of GA is the presence of double refraction, needle-shaped urate crystals found in the synovial fluid of the joint when observed under a polarizing microscope; however, clinicians are often reluctant to perform a procedure as invasive as joint puncture; therefore, imaging procedures have become necessary for proper clinical diagnosis. High-frequency ultrasound (US) is widely used to detect joint effusion, synovial hyperplasia and bone and articular soft tissue lesions in GA. The main typical US characteristics of GA are the double-contour sign of the articular cartilage, snowstorm appearance of the joint fluid and hyperechoes around the periarticular tendon (Chowalloor et al. 2013; Leng et al. 2014); however, it is highly challenging to distinguish GA from non-GA when the sonography results are not typical, such as in the early stage of the disease. At this time, the main typical US
INTRODUCTION Gout is a disease caused by a purine metabolism disorder and a reduction in uric acid excretion that is induced by either genetic or acquired factors. The clinical characteristics of gout include hyperuricemia, joint deformities, chronic gouty arthritis (GA), recurrent acute monoarthritis and uric acid sodium salt (sodium urate)-formed deposits of arthritic calculus. Severe cases can cause bone and joint lesions and joint disorders and deformities. In recent years, the prevalence of gout has gradually Address correspondence to: Li Qiu, Department of Ultrasound, West China Hospital of Sichuan University, No. 37 Guo Xue Xiang, Chengdu 610041, Sichuan, China. E-mail: [email protected] Conflict of interest disclosure: Yuanjiao Tang, Feng Yan, Yujia Yang, Xi Xiang, Liyun Wang, Lingyan Zhang and Li Qiu declare that: We have no proprietary, financial, professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the article entitled: Value of Shear Wave Elastography in the Diagnosis of Gouty Arthritis and Non-gouty Arthritis. 1
2
Ultrasound in Medicine and Biology
characteristics of GA cannot be detected, and the ultrasonic manifestations of GA only include only joint effusion and synovial hyperplasia, which can be detected in any kind of arthritis. Ultrasound elastography is a non-invasive exploration method that creates an image of the mechanical characteristics of the tissue—elasticity and stiffness (Bamber et al. 2013). Traditional strain elastography has several disadvantages, including unstable image quality, substantial inter-observer variability and the inability to provide a quantitative assessment of local stiffness (Chang et al. 2011; Zhou et al. 2014). Shear wave elastography (SWE) is a new technique in which several pushing focused beams are transmitted at increasing depths to generate a quasi-plane SW front. After generation of SWs, successive raw radiofrequency information is acquired over the complete region of interest (ROI) at a very high frame rate using ultrafast imaging equipment. Quantitative information is derived as SW velocity (in m/s) or Young’s elastic modulus (in kPa) (Vlad et al. 2015). This imaging method is reproducible, quantitative and operator independent. SWE has been applied in tissues, such as breast, thyroid, liver, kidney, muscle and tendon (Bob et al. 2015; Chen et al. 2013; Dillman et al. 2015; Vlad et al. 2015; Zhou et al. 2014), where it proved to be successful for the detection of tissue changes related to diseases or physiologic processes that might not have been detected using traditional US alone; however, to our knowledge, there have been no reports on its application for differentiating GA from non-GA.
METHODS Patients This retrospective study comprised 39 patients with GA and 55 patients with non-GA from the Department of Ultrasound at West China Hospital of Sichuan University from January to September 2015. The study was approved by the ethics committees of the West China Hospital of Sichuan University, and all patients provided informed consent for scientific analysis of their imaging data at the time of their examinations. Patients with GA or non-GA were randomly selected from outpatient and inpatient services. After the conventional US evaluation, an elastography procedure was performed on each patient. The GA group comprised 52 joints with chronic GA and 31 with an acute attack of chronic GA. Inclusion criteria for the GA group were compliance with the diagnostic criteria of the American Rheumatism Association guidelines; presence of at least one joint that currently has or once had clinical manifestations of acute gout; and hyperuricemia. The exclusion
Volume -, Number -, 2017
criteria were pregnancy; poor inspection compliance; and presence of other joint diseases. The non-GA group comprised 36 joints with rheumatoid arthritis, 20 with undifferentiated arthritis, 11 with psoriatic arthritis, 8 with tuberculous arthritis, 8 with osteoarthritis, 5 with arthritis caused by adultonset Still’s disease, 3 with traumatic arthritis and 2 with arthritis caused by ankylosing spondylitis. Criteria for inclusion in the non-GA group were compliance with the diagnostic criteria of the American Rheumatism Association guidelines or pathologic diagnosis; and normal serum uric acid concentration. The exclusion criteria were pregnancy; poor inspection compliance; and presence of two or more joint diseases. Table 1 summarizes the main characteristics of the study patients. Ultrasound and shear wave elastography examination Joints were first evaluated using conventional US (IU22, Philips, Amsterdam, Netherlands). The probe frequency was 8.0–12 MHz, and the musculoskeletal condition was selected. Based on the echo intensity of the joint lesion, as evaluated with conventional US (joint effusion was excluded), the GA group was divided into three subgroups: hypo-echoic, slightly hyper-echoic and hyperechoic. There were 46 joints in hypo-echoic GA subgroup, 20 in the slightly hyper-echoic GA subgroup and 17 in the hyper-echoic GA subgroup. In the first subgroup, the lesions were hypo-echoic in the joint capsule, with or without punctiform hyper-echoic images seen. In the second subgroup, scattered punctate hyper-echoes were seen in the joint cavity with or without slight acoustic shadows. In the third subgroup, multiple hyper-echoes were seen in the joint capsules with obvious acoustic shadows (Fig. 1). The joints were classified into the first subgroup if they had either both hypo-echoic and slight hyper-echoic lesions, or hypo-echoic and hyper-echoic lesions or all three. Those who had both slightly hyperechoic and hyper-echoic lesions were classified into the second subgroup. During US examination, an intraarticular color Doppler signal was graded from 0 to 3 (0 5 absence, no intra-articular flow; 1 5 mild, singlevessel signal or isolated signals; 2 5 moderate, confluent vessels; 3 5 marked, vessel signals in more than one-half of the intra-articular area) (Fig. 2) (Naredo et al. 2007). Shear wave elastography was performed on the joint lesions (joint effusion was excluded) with the Aixplorer system (SuperSonic Imagine, Aix-en-Provence, France) after conventional US examination, using a 15- to 4-MHz high-resolution linear transducer with the musculoskeletal preset. To undergo this method, patients had to be able to keep the joints stable when the examination was done; patients who could not keep the joints stable when the examination was done, such as those with
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
3
Table 1. Patient characteristics
Parameter Number of patients Age (y) Range Mean 6 SD Male Body mass index (kg/mm2) Number of joints All joints
Gouty arthritis group
Non-gouty arthritis group
39
55
29–72 49.44 6 11.80 36 (86.15%) 24.48 6 2.94 83 Metatarsophalangeal (n 5 35) Wrist (n 5 12) Knee (n 5 12) Ankle (n 5 12) Interphalangeal (n 5 5) Tarsal (n 5 5) Metacarpophalangeal (n 5 2)
9–76 45.36 6 17.26 26 (48.15%) 22.50 6 3.68 93 Wrist (n 5 30) Knee (n 5 28) Metatarsophalangeal (n 5 8) Ankle (n 5 8) Metacarpophalangeal (n 5 6) Elbow (n 5 6) Interphalangeal (n 5 3) Hip (n 5 2) Tarsal (n 5 2)
Parkinson’s and mental disorders and physically frail patients, were excluded. Patients whose joint lesions include only joint effusion or intra-articular free bodies should also be excluded. The SWE examination was performed on the hypo-echoic lesion if the joint had both hypo-echoic and slightly hyper-echoic lesions, or both hypo-echoic and hyper-echoic lesions or all three types of lesions, and was performed on the slightly hyperechoic lesion if the joint had both slightly hyper-echoic and hyper-echoic lesions. The SWE examination was performed on lesions with the higher Doppler signal grade if one joint had more than two grades. The ultrasound transducer was positioned on the longitudinal section of the joints. The transducer was held in place with minimal pressure on the skin and kept immobile while the SWE images were obtained. Shear wave elastography images were displayed together with the gray-scale US picture. A square ROI was preset for obtaining the SWE images and was marked on the lower-echo-intensity lesions on the gray scale. If the joint had both hypo-echoic and slightly hyperechoic lesions, or both hypo-echoic and hyper-echoic lesions or all three types of lesions, we marked the ROI on the hypo-echoic lesion. And if the joint had both slightly hyper-echoic and hyper-echoic lesions, we marked the ROI on the slightly hyper-echoic lesion. In the meantime,
visible bony fragments and effusion of the lesions were avoided. In the ROI, tissue stiffness was delineated using color mapping, in which very soft tissues were coded in dark blue, with areas of increasing stiffness coded in light blue, green, orange and red. The maximum elastic modulus (Emax), mean elastic modulus (Emean), minimum elastic modulus (Emin) and elastic modulus standard deviation (ESD) were automatically calculated by the US system (Fig. 3). Measurements were performed three times on each joint, and the average of the three values was calculated for each parameter. The US and SWE examinations were performed by two different radiologists with 5.0 and 10 y of experience in musculoskeletal US, respectively. Statistical analyses SPSS Version 19.0 (IBM, Armonk, NY, USA) was used for data analysis, and p , 0.05 was the significance level. Color Doppler signal grades were compared between GA and non-GA groups using the Mann–Whitney U-test. Emin, Emean, Emax and ESD of the GA and non-GA groups were compared using an independent t-test. For the non-GA group and the hypo-echoic GA, slightly hyper-echoic GA and hyper-echoic GA subgroups, Emin, Emean, Emax, and ESD were compared both among
Fig. 1. Sonography of metatarsophalangeal joints in (a) hypo-echoic subgroup, (b) slightly hyper-echoic subgroup and (c) hyper-echoic subgroup. The white arrows indicate the different echo intensities of the joint lesion in the three groups.
4
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 2. Intra-articular color Doppler signal grades of joint lesions. (a) Grade 0 joint lesion in a metatarsophalangeal joint. (b) Grade 1 joint lesion in a metacarpophalangeal joint. (c) Grade 2 joint lesion in a wrist joint. (d) Grade 3 joint lesion in a knee joint.
the groups/subgroups and between two groups/subgroups using the Kruskal–Wallis rank test and Mann–Whitney U-test with a Bonferroni correction to p value, respectively. The diagnostic performance of SWE for GA was evaluated using the receiver operating characteristic (ROC) analysis based on the data from the non-GA group and the hypo-echoic GA subgroup. According to Youden’s method, the value that provided the maximum sum of sensitivity (Se) and specificity (Sp) was considered the best diagnostic cutoff value. The positive likelihood ratio (1LR) and negative likelihood ratio (–LR) were also calculated. RESULTS Comparison of color Doppler signal grades between GA and non-GA groups The difference in color Doppler signal grades between the GA and non-GA groups was not significant (p . 0.05) (Table 2). Diagnostic performances of quantitative SWE features Emin, Emean, Emax and ESD were significantly higher in the GA group than in the non-GA group (p , 0.01 for all) (Fig. 4) and were highest in the hyper-echoic GA subgroup; the differences were significant (p # 0.001 for all). Emin, Emean and Emax were significantly higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD
was significantly higher in the slightly hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all) (Fig. 5). Results of ROC analysis The results of the ROC analysis are illustrated in Figure 6 and outlined in Table 3. For Emin, the area under the curve (AUC) was 0.749. The best cutoff value for predicting GA was 29.40 kPa, with an Se and an Sp of 76.1% and 65.6%, respectively, a 1 LR of 2.21 and a –LR of 0.36. For Emean, the AUC was 0.877. The best cutoff value was 45.35 kPa, with an Se of 87.0% and an Sp of 75.3% for the prediction of GA. In this case, the 1LR and –LR were 3.52 and 0.17, respectively. For Emax, the AUC was 0.896, the highest of the four parameters. The best cutoff value for predicting GA was 67.54 kPa with an Se of 87.0%, an Sp of 87.1%, a 1 LR of 6.74 and a –LR of 0.15. For ESD, the AUC was 0.886. The best cutoff value for predicting the GA was 7.85 kPa with an Se of 82.6%, an Sp of 77.4%, a 1 LR of 3.65 and a –LR of 0.22. DISCUSSION Gout is a heterogeneous disease caused by increased serum uric acid that is induced by disorders affecting
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
5
Fig. 3. Shear wave elastography (SWE) images of joint lesions in the four subgroups. (a) Metatarsophalangeal joint in non-gouty arthritis group. (b) Metatarsophalangeal joint in hypo-echoic gouty arthritis subgroup. (c) Knee joint in slightly hyper-echoic gouty arthritis subgroup. (d) Metatarsophalangeal joint in hyper-echoic gouty arthritis subgroup.
purine metabolism and those that cause an inability to excrete uric acid. The disease is often the primary manifestation of the onset of arthritis. Uric acid salts, or urates, precipitate in the synovium or synovial cavity as microcrystals, and excessive urate deposits are the basis for an acute inflammatory response, which leads to tophus formation (Chilappa et al. 2010; Smith et al. 2011; Tausche et al. 2009). Several studies have confirmed that early diagnosis and treatment of GA can Table 2. Comparison of color Doppler signal grades between GA and non-GA groups Color Doppler signal grade 0 1 2 3
GA group (n 5 83)
Non-GA (n 5 93)
p
37 (44.6%) 21 (25.3%) 18 (21.7%) 7 (8.4%)
39 (41.9%) 29 (31.2%) 15 (16.1%) 10 (10.8%)
.0.05
GA 5 gouty arthritis.
significantly improve the prognosis; therefore, the symptoms of GA should not be ignored. Compared with radiography, US is relatively easy to perform, emits no radiation and is not cost prohibitive, which is why it has been widely used in the diagnosis of GA; however, arthritis that is caused by different diseases could have similar US manifestations, such as joint effusion, synovial hyperplasia and bone and articular soft tissue lesions, when the US manifestations are not typical. This study was aimed at determining whether there is any value to using SWE in the diagnosis of GA compared with non-GA, particularly when the US manifestations are not typical. We are not aware of whether the intra-articular color Doppler signal has any influence on the SWE results; however, in our study, the difference in color Doppler signal grades between the GA and non-GA groups was not significant.
6
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 4. Box-and-whisker plots of Emax, Emean, Emin and ESD in joint lesions of gouty arthritis (GA) group and non-gouty arthritis (non-GA) group. The top and bottom of each box represent the 75th and 25th percentiles, respectively; the horizontal line in each box represents the median; and the top and bottom of the whiskers represent the minimum and maximum, respectively. *p , 0.05. Emax 5 elastic modulus, Emean 5 mean elastic modulus, Emin 5 minimum elastic modulus, ESD 5 elastic modulus standard deviation.
Emin, Emean, Emax and ESD were higher in the GA group than in non-GA group, and the differences were significant (p , 0.01 for all). We speculate that the urate precipitate in the synovium or synovial cavity in GA makes joint lesions harder and that the joint lesions in the GA group were more heterogeneous, which made ESD higher than in the non-GA group. On the basis of our study, of all the groups and subgroups, Emin, Emean, Emax and ESD were highest in the hyper-echoic GA subgroup, and the differences were sig-
nificant (p # 0.001 for all). In the hyper-echoic GA subgroup, multiple hyper-echoes were seen in the joint capsules with obvious acoustic shadows. This sign indicated the presence of intra-articular urate crystal precipitation and tophi, some of which had internal calcification (de Avila et al. 2011); therefore, the joint lesions in the hyper-echoic GA subgroup were the hardest of all the groups and subgroups, which accounted for the higher Emin, Emean and Emax. Because the joint lesions in the hyper-echoic GA subgroup contained several different
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
7
Fig. 5. Box-and-whisker plots of Emin, Emean, Emax and ESD in joint lesions of non-gouty arthritis (non-GA) group, hypoechoic GA subgroup, slightly hyper-echoic GA subgroup and hyper-echoic GA subgroup. The top and bottom of each box represent the 75th and 25th percentiles, respectively; the horizontal line in each box represents the median; and the top and bottom of the whiskers represent the minimum and maximum, respectively. *p , 0.001, versus non-GA group. # p , 0.001, versus GA-1 group. $p 5 0.001, versus GA-1 group. &p , 0.001 versus GA-2 group. GA-1 5 hypo-echoic GA subgroup; GA-2 5 slightly hyper-echoic GA subgroup; GA-3 5 hyper-echoic GA subgroup; Emax 5 elastic modulus; Emean 5 mean elastic modulus; Emin 5 minimum elastic modulus; ESD 5 elastic modulus standard deviation.
ingredients, ESD was higher in this subgroup than in the other group/subgroups. In our study, Emin, Emean and Emax were significantly higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD was significantly higher in the slightly hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). In the slightly hyper-echoic GA subgroup, scattered punctate hyperechoes were seen in the joint cavity with or without slight acoustic shadows. This sign indicates the formation of urate crystals in the joint cavity, some becoming small tophi. This was supported by our results in that the joint lesions in the slightly hyper-echoic GA subgroup were harder than those in the hypo-echoic GA subgroup and non-GA group and more heterogeneous than those in the non-GA group. ESD was higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup, but the difference was not significant. The reason might have been that our sample size was not large enough. In a subsequent study, we will increase the sample size and determine the effect on the
difference in ESD between the slightly hyper-echoic GA subgroup and the hypo-echoic GA subgroup. On the basis of our study, Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all). In the hypo-echoic GA subgroup, the lesions were hypoechoic in the joint capsule, with or without punctiform hyper-echoes in the images. This sign indicated that synovial hyperplasia with a bit of urate crystal precipitation made the joint lesions harder and more heterogeneous than those in the non-GA group and resulted in higher Emin, Emean, Emax and ESD values in the hypo-echoic GA subgroup than in the non-GA group. When the US manifestations are not typical of GA, as in the early stages of GA, these manifestations tend to include only joint effusion and synovial hyperplasia, and the typical US characteristics, such as the doublecontour sign of articular cartilage, snowstorm appearance of the joint fluid and hyper-echo around the periarticular tendon, are rarely seen. Because non-GA produces similar US manifestations, it is difficult to distinguish
8
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 6. Receiver operating characteristic curve for the prediction of gouty arthritis using shear wave elastography.
GA from non-GA using only this method. The manifestation of synovial hyperplasia tends to be hypo-echoic in the joint capsule; therefore, it is important to determine whether the elastic parameters of the hypo-echoic GA subgroup and non-GA group different on the SWE images. Our study indicated that the GA group had higher Emin, Emean, Emax and ESD values. To ensure early treatment, this is extremely important in the differential diagnosis of GA and non-GA when the US manifestations are not typical.
In the slightly hyper-echoic GA and hyper-echoic GA subgroups, it was common to see the typical US characteristics, such as tophi in the joint capsule and snowstorm appearance of the joint fluid, which made it easier to differentially diagnose GA from non-GA; therefore, it was important to determine the diagnostic value of SWE in the hypo-echoic GA subgroup and non-GA group, as previously mentioned. To that end, we determined the ROC curve based on the data from the hypoechoic GA subgroup and the non-GA group.
Table 3. Diagnostic performance of shear wave elastography features Parameter
AUC
Cutoff value (kPa)
Sensitivity (%)
Specificity (%)
1LR
–LR
Emin Emean Emax ESD
0.749 0.877 0.896 0.886
29.40 45.35 67.54 7.85
76.1 87.0 87.0 82.6
65.6 75.3 87.1 77.4
2.21 3.52 6.74 3.65
0.36 0.17 0.15 0.22
AUC 5 area under the curve; 1LR 5 positive likelihood ratio; –LR 5 negative likelihood ratio; Emax 5 elastic modulus; Emean 5 mean elastic modulus; Emin 5 minimum elastic modulus; ESD, 5 elastic modulus standard deviation.
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
According to our data, the AUCs for Emin, Emean, Emax and ESD, were all .0.7; the AUC of Emax was largest, with the highest 1LR, Se and Sp and the lowest –LR of the four cutoff values in this selected population. Previous studies (Grgurevic et al. 2015; Shi et al. 2015) had reported that SWE can accurately differentiate between the stages of liver fibrosis and that combining SWE parameters can improve the accuracy of US in the differentiation of benign from malignant breast lesions. Our results prove that SWE can differentiate GA with hypoechoic joint lesions and non-GA for all elastic parameters. Our results also indicate that Emax was the bestperforming quantitative SWE feature, which is in agreement with Wang et al. (2013), who determined that Emax was the best-performing quantitative SWE feature in the diagnosis of breast tumors. The best cutoff values for predicting GA were 29.4 kPa for Emin, 45.35 kPa for Emean, 67.54 kPa for Emax and 7.85 kPa for ESD. These results meant that clinicians can use SWE to differentially diagnose GA from non-GA when the US manifestations are not typical. There were some limitations to this study. First, our GA group did not include acute GA. Second, we did not compare SWE with conventional US features. Perhaps SWE combined with conventional US can improve the Se and Sp in the differential diagnosis of GA and nonGA. Third, the non-GA group did not include other crystal-induced arthritis, such as calcium pyrophosphate deposition disease (CPPD). These limitations should be addressed in future work. CONCLUSIONS Our study revealed that SWE could be used to quantitatively differentiate GA and non-GA. The joint lesions in GA are significantly stiffer than those in non-GA. Among other elastic modulus values, Emax appeared to be the most discriminatory parameter in the diagnosis of GA from non-GA. These findings could be particularly useful when joint lesions in GA are hypo-echoic without typical US manifestations. This means that in the future, by using SWE, clinicians can diagnose GA earlier noninvasively and offer the patients early treatment. In our future study, we will increase sample sizes and pay more attention to the value of SWE in the diagnosis of different kinds of arthritis. Acknowledgments—This study was supported by grants from National Natural Science Foundation of China (81671696), Sichuan Science
9
and Technology Plan (2015 SZ0125) and Chengdu Science and Technology Plan (2014-HM01-00176-SF).
REFERENCES Bamber J, Cosgrove D, Dietrich CF, Fromageau J, Bojunga J, Calliada F, Cantisani V, Correas JM, D’Onofrio M, Drakonaki EE, Fink M, Friedrich-Rust M, Gilja OH, Havre RF, Jenssen C, Klauser AS, Ohlinger R, Saftoiu A, Schaefer F, Sporea I, Piscaglia F. EFSUMB guidelines and recommendations on the clinical use of ultrasound elastography: Part 1. Basic principles and technology. Ultraschall Med 2013;34:169–184. Bob F, Bota S, Sporea I, Sirli R, Popescu A, Schiller A. Relationship between the estimated glomerular filtration rate and kidney shear wave speed values assessed by acoustic radiation force impulse elastography: A pilot study. J Ultrasound Med 2015;34:649–654. Chang JM, Moon WK, Cho N, Kim SJ. Breast mass evaluation: Factors influencing the quality of US elastography. Radiology 2011;259: 59–64. Chen XM, Cui LG, He P, Shen WW, Qian YJ, Wang JR. Shear wave elastographic characterization of normal and torn achilles tendons. J Ultrasound Med 2013;32:449–455. Chilappa CS, Aronow WS, Shapiro D, Sperber K, Patel U, Ash JY. Gout and hyperuricemia. Compr Ther 2010;36:3–13. Chowalloor PV, Keen HI. A systematic review of ultrasonography in gout and asymptomatic hyperuricaemia. Ann Rheum Dis 2013;72: 638–645. de Avila Fernandes E, Kubota ES, Sandim GB, Mitraud SA, Ferrari AJ, Fernandes AR. Ultrasound features of tophi in chronic tophaceous gout. Skeletal Radiol 2011;40:309–315. Dillman JR, Heider A, Bilhartz JL, Smith EA, Keshavarzi N, Rubin JM, Lopez MJ. Ultrasound shear wave speed measurements correlate with liver fibrosis in children. Pediatr Radiol 2015;45:1480–1488. Grgurevic I, Puljiz Z, Brnic D, Bokun T, Heinzl R, Lukic A, Luksic B, Kujundzic M, Brkljacic B. Liver and spleen stiffness and their ratio assessed by real-time two dimensional-shear wave elastography in patients with liver fibrosis and cirrhosis due to chronic viral hepatitis. Eur Radiol 2015;25:3214–3221. Leng QY, Tang YJ, Zhang LY, Xiang X, Qi L, Su BH. Diagnostic value of ultrasound imaging in chronic gouty arthritis. Sichuan Da Xue Bao Yi Xue Ban 2014;45:424–427. 446. Naredo E, Collado P, Cruz A, Palop MJ, Cabero F, Richi P, Carmona L, Crespo M. Longitudinal power Doppler ultrasonographic assessment of joint inflammatory activity in early rheumatoid arthritis: Predictive value in disease activity and radiologic progression. Arthritis Rheum 2007;57:116–124. Shi XQ, Li JL, Wan WB, Huang Y. A set of shear wave elastography quantitative parameters combined with ultrasound BI-RADS to assess benign and malignant breast lesions. Ultrasound Med Biol 2015;41:960–966. Smith HS, Bracken D, Smith JM. Gout: Current insights and future perspectives. J Pain 2011;12:1113–1129. Tausche AK, Jansen TL, Schr€oder HE, Bornstein SR, Aringer M, M€uller-Ladner U. Gout—Current diagnosis and treatment. Dtsch Arztebl Int 2009;106:549–555. Vlad M, Golu I, Bota S, Vlad A, Timar B, Timar R, Sporea I. Real-time shear wave elastography may predict autoimmune thyroid disease. Wien Klin Wochenschr 2015;127:330–336. Wang ZL, Li JL, Li M, Huang Y, Wan WB, Tang J. Study of quantitative elastography with supersonic shear imaging in the diagnosis of breast tumours. Radiol Med (Torino) 2013;118:583–590. Zhou J, Zhan W, Chang C, Zhang X, Jia Y, Dong Y, Zhou C, Sun J, Grant EG. Breast Lesions: Evaluation with shear wave elastography, with special emphasis on the ‘‘stiff rim’’ sign. Radiology 2014;272: 63–72.
http://dx.doi.org/10.1016/j.ultrasmedbio.2016.12.012
d
Original Contribution VALUE OF SHEAR WAVE ELASTOGRAPHY IN THE DIAGNOSIS OF GOUTY AND NON-GOUTY ARTHRITIS YUANJIAO TANG,* FENG YAN,*y YUJIA YANG,* XI XIANG,* LIYUN WANG,* LINGYAN ZHANG,* and LI QIU* * Department of Ultrasound, West China Hospital of Sichuan University, Chengdu, Sichuan, China; and y Clinical Ultrasound Imaging Drug Research Lab, West China Hospital of Sichuan University, Chengdu, Sichuan, China (Received 22 August 2016; revised 15 December 2016; in final form 20 December 2016)
Abstract—Our aim was to analyze the diagnostic performance of shear wave elastography (SWE) in the diagnosis of gouty arthritis (GA) and non-gouty arthritis (non-GA). Thirty-nine patients in the GA group and 55 patients in the non-GA group were included in the study. Based on the echo intensity of the joint lesions, the GA group was subdivided into hypo-echoic GA, slightly hyper-echoic GA and hyper-echoic GA subgroups. Quantitative SWE features were evaluated and receiver operating characteristic analysis was performed. On the basis of the study, the elastic modulus (Emax), mean elastic modulus (Emean), minimum elastic modulus (Emin) and elastic modulus standard deviation (ESD) were significantly higher in the GA group than in the non-GA group and were highest in the hyper-echoic GA subgroup (p , 0.01 for all). Emin, Emean and Emax were significantly higher in the hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD was significantly higher in the hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all). Based on the hypo-echoic GA subgroup and non-GA group, areas under the receiver operating characteristic curves for the prediction of GA were 0.749 for Emin, 0.877 for Emean, 0.896 for Emax and 0.886 for ESD, with optimal cutoff values of 29.40 kPa for Emin, 45.35 kPa for Emean, 67.54 kPa for Emax and 7.85 kPa for ESD. Our results indicate that SWE can differentially diagnose GA and non-GA, especially when the ultrasound manifestations are not typical. (E-mail: [email protected]) Ó 2016 World Federation for Ultrasound in Medicine & Biology. Key Words: Gouty arthritis, Non-gouty arthritis, Ultrasound, Shear wave elastography, Stiffness.
increased with improved living standards (Chilappa et al. 2010). The most reliable diagnostic criterion of GA is the presence of double refraction, needle-shaped urate crystals found in the synovial fluid of the joint when observed under a polarizing microscope; however, clinicians are often reluctant to perform a procedure as invasive as joint puncture; therefore, imaging procedures have become necessary for proper clinical diagnosis. High-frequency ultrasound (US) is widely used to detect joint effusion, synovial hyperplasia and bone and articular soft tissue lesions in GA. The main typical US characteristics of GA are the double-contour sign of the articular cartilage, snowstorm appearance of the joint fluid and hyperechoes around the periarticular tendon (Chowalloor et al. 2013; Leng et al. 2014); however, it is highly challenging to distinguish GA from non-GA when the sonography results are not typical, such as in the early stage of the disease. At this time, the main typical US
INTRODUCTION Gout is a disease caused by a purine metabolism disorder and a reduction in uric acid excretion that is induced by either genetic or acquired factors. The clinical characteristics of gout include hyperuricemia, joint deformities, chronic gouty arthritis (GA), recurrent acute monoarthritis and uric acid sodium salt (sodium urate)-formed deposits of arthritic calculus. Severe cases can cause bone and joint lesions and joint disorders and deformities. In recent years, the prevalence of gout has gradually Address correspondence to: Li Qiu, Department of Ultrasound, West China Hospital of Sichuan University, No. 37 Guo Xue Xiang, Chengdu 610041, Sichuan, China. E-mail: [email protected] Conflict of interest disclosure: Yuanjiao Tang, Feng Yan, Yujia Yang, Xi Xiang, Liyun Wang, Lingyan Zhang and Li Qiu declare that: We have no proprietary, financial, professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the article entitled: Value of Shear Wave Elastography in the Diagnosis of Gouty Arthritis and Non-gouty Arthritis. 1
2
Ultrasound in Medicine and Biology
characteristics of GA cannot be detected, and the ultrasonic manifestations of GA only include only joint effusion and synovial hyperplasia, which can be detected in any kind of arthritis. Ultrasound elastography is a non-invasive exploration method that creates an image of the mechanical characteristics of the tissue—elasticity and stiffness (Bamber et al. 2013). Traditional strain elastography has several disadvantages, including unstable image quality, substantial inter-observer variability and the inability to provide a quantitative assessment of local stiffness (Chang et al. 2011; Zhou et al. 2014). Shear wave elastography (SWE) is a new technique in which several pushing focused beams are transmitted at increasing depths to generate a quasi-plane SW front. After generation of SWs, successive raw radiofrequency information is acquired over the complete region of interest (ROI) at a very high frame rate using ultrafast imaging equipment. Quantitative information is derived as SW velocity (in m/s) or Young’s elastic modulus (in kPa) (Vlad et al. 2015). This imaging method is reproducible, quantitative and operator independent. SWE has been applied in tissues, such as breast, thyroid, liver, kidney, muscle and tendon (Bob et al. 2015; Chen et al. 2013; Dillman et al. 2015; Vlad et al. 2015; Zhou et al. 2014), where it proved to be successful for the detection of tissue changes related to diseases or physiologic processes that might not have been detected using traditional US alone; however, to our knowledge, there have been no reports on its application for differentiating GA from non-GA.
METHODS Patients This retrospective study comprised 39 patients with GA and 55 patients with non-GA from the Department of Ultrasound at West China Hospital of Sichuan University from January to September 2015. The study was approved by the ethics committees of the West China Hospital of Sichuan University, and all patients provided informed consent for scientific analysis of their imaging data at the time of their examinations. Patients with GA or non-GA were randomly selected from outpatient and inpatient services. After the conventional US evaluation, an elastography procedure was performed on each patient. The GA group comprised 52 joints with chronic GA and 31 with an acute attack of chronic GA. Inclusion criteria for the GA group were compliance with the diagnostic criteria of the American Rheumatism Association guidelines; presence of at least one joint that currently has or once had clinical manifestations of acute gout; and hyperuricemia. The exclusion
Volume -, Number -, 2017
criteria were pregnancy; poor inspection compliance; and presence of other joint diseases. The non-GA group comprised 36 joints with rheumatoid arthritis, 20 with undifferentiated arthritis, 11 with psoriatic arthritis, 8 with tuberculous arthritis, 8 with osteoarthritis, 5 with arthritis caused by adultonset Still’s disease, 3 with traumatic arthritis and 2 with arthritis caused by ankylosing spondylitis. Criteria for inclusion in the non-GA group were compliance with the diagnostic criteria of the American Rheumatism Association guidelines or pathologic diagnosis; and normal serum uric acid concentration. The exclusion criteria were pregnancy; poor inspection compliance; and presence of two or more joint diseases. Table 1 summarizes the main characteristics of the study patients. Ultrasound and shear wave elastography examination Joints were first evaluated using conventional US (IU22, Philips, Amsterdam, Netherlands). The probe frequency was 8.0–12 MHz, and the musculoskeletal condition was selected. Based on the echo intensity of the joint lesion, as evaluated with conventional US (joint effusion was excluded), the GA group was divided into three subgroups: hypo-echoic, slightly hyper-echoic and hyperechoic. There were 46 joints in hypo-echoic GA subgroup, 20 in the slightly hyper-echoic GA subgroup and 17 in the hyper-echoic GA subgroup. In the first subgroup, the lesions were hypo-echoic in the joint capsule, with or without punctiform hyper-echoic images seen. In the second subgroup, scattered punctate hyper-echoes were seen in the joint cavity with or without slight acoustic shadows. In the third subgroup, multiple hyper-echoes were seen in the joint capsules with obvious acoustic shadows (Fig. 1). The joints were classified into the first subgroup if they had either both hypo-echoic and slight hyper-echoic lesions, or hypo-echoic and hyper-echoic lesions or all three. Those who had both slightly hyperechoic and hyper-echoic lesions were classified into the second subgroup. During US examination, an intraarticular color Doppler signal was graded from 0 to 3 (0 5 absence, no intra-articular flow; 1 5 mild, singlevessel signal or isolated signals; 2 5 moderate, confluent vessels; 3 5 marked, vessel signals in more than one-half of the intra-articular area) (Fig. 2) (Naredo et al. 2007). Shear wave elastography was performed on the joint lesions (joint effusion was excluded) with the Aixplorer system (SuperSonic Imagine, Aix-en-Provence, France) after conventional US examination, using a 15- to 4-MHz high-resolution linear transducer with the musculoskeletal preset. To undergo this method, patients had to be able to keep the joints stable when the examination was done; patients who could not keep the joints stable when the examination was done, such as those with
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
3
Table 1. Patient characteristics
Parameter Number of patients Age (y) Range Mean 6 SD Male Body mass index (kg/mm2) Number of joints All joints
Gouty arthritis group
Non-gouty arthritis group
39
55
29–72 49.44 6 11.80 36 (86.15%) 24.48 6 2.94 83 Metatarsophalangeal (n 5 35) Wrist (n 5 12) Knee (n 5 12) Ankle (n 5 12) Interphalangeal (n 5 5) Tarsal (n 5 5) Metacarpophalangeal (n 5 2)
9–76 45.36 6 17.26 26 (48.15%) 22.50 6 3.68 93 Wrist (n 5 30) Knee (n 5 28) Metatarsophalangeal (n 5 8) Ankle (n 5 8) Metacarpophalangeal (n 5 6) Elbow (n 5 6) Interphalangeal (n 5 3) Hip (n 5 2) Tarsal (n 5 2)
Parkinson’s and mental disorders and physically frail patients, were excluded. Patients whose joint lesions include only joint effusion or intra-articular free bodies should also be excluded. The SWE examination was performed on the hypo-echoic lesion if the joint had both hypo-echoic and slightly hyper-echoic lesions, or both hypo-echoic and hyper-echoic lesions or all three types of lesions, and was performed on the slightly hyperechoic lesion if the joint had both slightly hyper-echoic and hyper-echoic lesions. The SWE examination was performed on lesions with the higher Doppler signal grade if one joint had more than two grades. The ultrasound transducer was positioned on the longitudinal section of the joints. The transducer was held in place with minimal pressure on the skin and kept immobile while the SWE images were obtained. Shear wave elastography images were displayed together with the gray-scale US picture. A square ROI was preset for obtaining the SWE images and was marked on the lower-echo-intensity lesions on the gray scale. If the joint had both hypo-echoic and slightly hyperechoic lesions, or both hypo-echoic and hyper-echoic lesions or all three types of lesions, we marked the ROI on the hypo-echoic lesion. And if the joint had both slightly hyper-echoic and hyper-echoic lesions, we marked the ROI on the slightly hyper-echoic lesion. In the meantime,
visible bony fragments and effusion of the lesions were avoided. In the ROI, tissue stiffness was delineated using color mapping, in which very soft tissues were coded in dark blue, with areas of increasing stiffness coded in light blue, green, orange and red. The maximum elastic modulus (Emax), mean elastic modulus (Emean), minimum elastic modulus (Emin) and elastic modulus standard deviation (ESD) were automatically calculated by the US system (Fig. 3). Measurements were performed three times on each joint, and the average of the three values was calculated for each parameter. The US and SWE examinations were performed by two different radiologists with 5.0 and 10 y of experience in musculoskeletal US, respectively. Statistical analyses SPSS Version 19.0 (IBM, Armonk, NY, USA) was used for data analysis, and p , 0.05 was the significance level. Color Doppler signal grades were compared between GA and non-GA groups using the Mann–Whitney U-test. Emin, Emean, Emax and ESD of the GA and non-GA groups were compared using an independent t-test. For the non-GA group and the hypo-echoic GA, slightly hyper-echoic GA and hyper-echoic GA subgroups, Emin, Emean, Emax, and ESD were compared both among
Fig. 1. Sonography of metatarsophalangeal joints in (a) hypo-echoic subgroup, (b) slightly hyper-echoic subgroup and (c) hyper-echoic subgroup. The white arrows indicate the different echo intensities of the joint lesion in the three groups.
4
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 2. Intra-articular color Doppler signal grades of joint lesions. (a) Grade 0 joint lesion in a metatarsophalangeal joint. (b) Grade 1 joint lesion in a metacarpophalangeal joint. (c) Grade 2 joint lesion in a wrist joint. (d) Grade 3 joint lesion in a knee joint.
the groups/subgroups and between two groups/subgroups using the Kruskal–Wallis rank test and Mann–Whitney U-test with a Bonferroni correction to p value, respectively. The diagnostic performance of SWE for GA was evaluated using the receiver operating characteristic (ROC) analysis based on the data from the non-GA group and the hypo-echoic GA subgroup. According to Youden’s method, the value that provided the maximum sum of sensitivity (Se) and specificity (Sp) was considered the best diagnostic cutoff value. The positive likelihood ratio (1LR) and negative likelihood ratio (–LR) were also calculated. RESULTS Comparison of color Doppler signal grades between GA and non-GA groups The difference in color Doppler signal grades between the GA and non-GA groups was not significant (p . 0.05) (Table 2). Diagnostic performances of quantitative SWE features Emin, Emean, Emax and ESD were significantly higher in the GA group than in the non-GA group (p , 0.01 for all) (Fig. 4) and were highest in the hyper-echoic GA subgroup; the differences were significant (p # 0.001 for all). Emin, Emean and Emax were significantly higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD
was significantly higher in the slightly hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all) (Fig. 5). Results of ROC analysis The results of the ROC analysis are illustrated in Figure 6 and outlined in Table 3. For Emin, the area under the curve (AUC) was 0.749. The best cutoff value for predicting GA was 29.40 kPa, with an Se and an Sp of 76.1% and 65.6%, respectively, a 1 LR of 2.21 and a –LR of 0.36. For Emean, the AUC was 0.877. The best cutoff value was 45.35 kPa, with an Se of 87.0% and an Sp of 75.3% for the prediction of GA. In this case, the 1LR and –LR were 3.52 and 0.17, respectively. For Emax, the AUC was 0.896, the highest of the four parameters. The best cutoff value for predicting GA was 67.54 kPa with an Se of 87.0%, an Sp of 87.1%, a 1 LR of 6.74 and a –LR of 0.15. For ESD, the AUC was 0.886. The best cutoff value for predicting the GA was 7.85 kPa with an Se of 82.6%, an Sp of 77.4%, a 1 LR of 3.65 and a –LR of 0.22. DISCUSSION Gout is a heterogeneous disease caused by increased serum uric acid that is induced by disorders affecting
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
5
Fig. 3. Shear wave elastography (SWE) images of joint lesions in the four subgroups. (a) Metatarsophalangeal joint in non-gouty arthritis group. (b) Metatarsophalangeal joint in hypo-echoic gouty arthritis subgroup. (c) Knee joint in slightly hyper-echoic gouty arthritis subgroup. (d) Metatarsophalangeal joint in hyper-echoic gouty arthritis subgroup.
purine metabolism and those that cause an inability to excrete uric acid. The disease is often the primary manifestation of the onset of arthritis. Uric acid salts, or urates, precipitate in the synovium or synovial cavity as microcrystals, and excessive urate deposits are the basis for an acute inflammatory response, which leads to tophus formation (Chilappa et al. 2010; Smith et al. 2011; Tausche et al. 2009). Several studies have confirmed that early diagnosis and treatment of GA can Table 2. Comparison of color Doppler signal grades between GA and non-GA groups Color Doppler signal grade 0 1 2 3
GA group (n 5 83)
Non-GA (n 5 93)
p
37 (44.6%) 21 (25.3%) 18 (21.7%) 7 (8.4%)
39 (41.9%) 29 (31.2%) 15 (16.1%) 10 (10.8%)
.0.05
GA 5 gouty arthritis.
significantly improve the prognosis; therefore, the symptoms of GA should not be ignored. Compared with radiography, US is relatively easy to perform, emits no radiation and is not cost prohibitive, which is why it has been widely used in the diagnosis of GA; however, arthritis that is caused by different diseases could have similar US manifestations, such as joint effusion, synovial hyperplasia and bone and articular soft tissue lesions, when the US manifestations are not typical. This study was aimed at determining whether there is any value to using SWE in the diagnosis of GA compared with non-GA, particularly when the US manifestations are not typical. We are not aware of whether the intra-articular color Doppler signal has any influence on the SWE results; however, in our study, the difference in color Doppler signal grades between the GA and non-GA groups was not significant.
6
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 4. Box-and-whisker plots of Emax, Emean, Emin and ESD in joint lesions of gouty arthritis (GA) group and non-gouty arthritis (non-GA) group. The top and bottom of each box represent the 75th and 25th percentiles, respectively; the horizontal line in each box represents the median; and the top and bottom of the whiskers represent the minimum and maximum, respectively. *p , 0.05. Emax 5 elastic modulus, Emean 5 mean elastic modulus, Emin 5 minimum elastic modulus, ESD 5 elastic modulus standard deviation.
Emin, Emean, Emax and ESD were higher in the GA group than in non-GA group, and the differences were significant (p , 0.01 for all). We speculate that the urate precipitate in the synovium or synovial cavity in GA makes joint lesions harder and that the joint lesions in the GA group were more heterogeneous, which made ESD higher than in the non-GA group. On the basis of our study, of all the groups and subgroups, Emin, Emean, Emax and ESD were highest in the hyper-echoic GA subgroup, and the differences were sig-
nificant (p # 0.001 for all). In the hyper-echoic GA subgroup, multiple hyper-echoes were seen in the joint capsules with obvious acoustic shadows. This sign indicated the presence of intra-articular urate crystal precipitation and tophi, some of which had internal calcification (de Avila et al. 2011); therefore, the joint lesions in the hyper-echoic GA subgroup were the hardest of all the groups and subgroups, which accounted for the higher Emin, Emean and Emax. Because the joint lesions in the hyper-echoic GA subgroup contained several different
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
7
Fig. 5. Box-and-whisker plots of Emin, Emean, Emax and ESD in joint lesions of non-gouty arthritis (non-GA) group, hypoechoic GA subgroup, slightly hyper-echoic GA subgroup and hyper-echoic GA subgroup. The top and bottom of each box represent the 75th and 25th percentiles, respectively; the horizontal line in each box represents the median; and the top and bottom of the whiskers represent the minimum and maximum, respectively. *p , 0.001, versus non-GA group. # p , 0.001, versus GA-1 group. $p 5 0.001, versus GA-1 group. &p , 0.001 versus GA-2 group. GA-1 5 hypo-echoic GA subgroup; GA-2 5 slightly hyper-echoic GA subgroup; GA-3 5 hyper-echoic GA subgroup; Emax 5 elastic modulus; Emean 5 mean elastic modulus; Emin 5 minimum elastic modulus; ESD 5 elastic modulus standard deviation.
ingredients, ESD was higher in this subgroup than in the other group/subgroups. In our study, Emin, Emean and Emax were significantly higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup and non-GA group (p , 0.001 for all), and ESD was significantly higher in the slightly hyper-echoic GA subgroup than in the non-GA group (p 5 0.001). In the slightly hyper-echoic GA subgroup, scattered punctate hyperechoes were seen in the joint cavity with or without slight acoustic shadows. This sign indicates the formation of urate crystals in the joint cavity, some becoming small tophi. This was supported by our results in that the joint lesions in the slightly hyper-echoic GA subgroup were harder than those in the hypo-echoic GA subgroup and non-GA group and more heterogeneous than those in the non-GA group. ESD was higher in the slightly hyper-echoic GA subgroup than in the hypo-echoic GA subgroup, but the difference was not significant. The reason might have been that our sample size was not large enough. In a subsequent study, we will increase the sample size and determine the effect on the
difference in ESD between the slightly hyper-echoic GA subgroup and the hypo-echoic GA subgroup. On the basis of our study, Emin, Emean, Emax and ESD were higher in the hypo-echoic GA subgroup than in the non-GA group, and the differences were significant (p , 0.001 for all). In the hypo-echoic GA subgroup, the lesions were hypoechoic in the joint capsule, with or without punctiform hyper-echoes in the images. This sign indicated that synovial hyperplasia with a bit of urate crystal precipitation made the joint lesions harder and more heterogeneous than those in the non-GA group and resulted in higher Emin, Emean, Emax and ESD values in the hypo-echoic GA subgroup than in the non-GA group. When the US manifestations are not typical of GA, as in the early stages of GA, these manifestations tend to include only joint effusion and synovial hyperplasia, and the typical US characteristics, such as the doublecontour sign of articular cartilage, snowstorm appearance of the joint fluid and hyper-echo around the periarticular tendon, are rarely seen. Because non-GA produces similar US manifestations, it is difficult to distinguish
8
Ultrasound in Medicine and Biology
Volume -, Number -, 2017
Fig. 6. Receiver operating characteristic curve for the prediction of gouty arthritis using shear wave elastography.
GA from non-GA using only this method. The manifestation of synovial hyperplasia tends to be hypo-echoic in the joint capsule; therefore, it is important to determine whether the elastic parameters of the hypo-echoic GA subgroup and non-GA group different on the SWE images. Our study indicated that the GA group had higher Emin, Emean, Emax and ESD values. To ensure early treatment, this is extremely important in the differential diagnosis of GA and non-GA when the US manifestations are not typical.
In the slightly hyper-echoic GA and hyper-echoic GA subgroups, it was common to see the typical US characteristics, such as tophi in the joint capsule and snowstorm appearance of the joint fluid, which made it easier to differentially diagnose GA from non-GA; therefore, it was important to determine the diagnostic value of SWE in the hypo-echoic GA subgroup and non-GA group, as previously mentioned. To that end, we determined the ROC curve based on the data from the hypoechoic GA subgroup and the non-GA group.
Table 3. Diagnostic performance of shear wave elastography features Parameter
AUC
Cutoff value (kPa)
Sensitivity (%)
Specificity (%)
1LR
–LR
Emin Emean Emax ESD
0.749 0.877 0.896 0.886
29.40 45.35 67.54 7.85
76.1 87.0 87.0 82.6
65.6 75.3 87.1 77.4
2.21 3.52 6.74 3.65
0.36 0.17 0.15 0.22
AUC 5 area under the curve; 1LR 5 positive likelihood ratio; –LR 5 negative likelihood ratio; Emax 5 elastic modulus; Emean 5 mean elastic modulus; Emin 5 minimum elastic modulus; ESD, 5 elastic modulus standard deviation.
SWE in diagnosis of gouty and non-gouty arthritis d Y. TANG et al.
According to our data, the AUCs for Emin, Emean, Emax and ESD, were all .0.7; the AUC of Emax was largest, with the highest 1LR, Se and Sp and the lowest –LR of the four cutoff values in this selected population. Previous studies (Grgurevic et al. 2015; Shi et al. 2015) had reported that SWE can accurately differentiate between the stages of liver fibrosis and that combining SWE parameters can improve the accuracy of US in the differentiation of benign from malignant breast lesions. Our results prove that SWE can differentiate GA with hypoechoic joint lesions and non-GA for all elastic parameters. Our results also indicate that Emax was the bestperforming quantitative SWE feature, which is in agreement with Wang et al. (2013), who determined that Emax was the best-performing quantitative SWE feature in the diagnosis of breast tumors. The best cutoff values for predicting GA were 29.4 kPa for Emin, 45.35 kPa for Emean, 67.54 kPa for Emax and 7.85 kPa for ESD. These results meant that clinicians can use SWE to differentially diagnose GA from non-GA when the US manifestations are not typical. There were some limitations to this study. First, our GA group did not include acute GA. Second, we did not compare SWE with conventional US features. Perhaps SWE combined with conventional US can improve the Se and Sp in the differential diagnosis of GA and nonGA. Third, the non-GA group did not include other crystal-induced arthritis, such as calcium pyrophosphate deposition disease (CPPD). These limitations should be addressed in future work. CONCLUSIONS Our study revealed that SWE could be used to quantitatively differentiate GA and non-GA. The joint lesions in GA are significantly stiffer than those in non-GA. Among other elastic modulus values, Emax appeared to be the most discriminatory parameter in the diagnosis of GA from non-GA. These findings could be particularly useful when joint lesions in GA are hypo-echoic without typical US manifestations. This means that in the future, by using SWE, clinicians can diagnose GA earlier noninvasively and offer the patients early treatment. In our future study, we will increase sample sizes and pay more attention to the value of SWE in the diagnosis of different kinds of arthritis. Acknowledgments—This study was supported by grants from National Natural Science Foundation of China (81671696), Sichuan Science
9
and Technology Plan (2015 SZ0125) and Chengdu Science and Technology Plan (2014-HM01-00176-SF).
REFERENCES Bamber J, Cosgrove D, Dietrich CF, Fromageau J, Bojunga J, Calliada F, Cantisani V, Correas JM, D’Onofrio M, Drakonaki EE, Fink M, Friedrich-Rust M, Gilja OH, Havre RF, Jenssen C, Klauser AS, Ohlinger R, Saftoiu A, Schaefer F, Sporea I, Piscaglia F. EFSUMB guidelines and recommendations on the clinical use of ultrasound elastography: Part 1. Basic principles and technology. Ultraschall Med 2013;34:169–184. Bob F, Bota S, Sporea I, Sirli R, Popescu A, Schiller A. Relationship between the estimated glomerular filtration rate and kidney shear wave speed values assessed by acoustic radiation force impulse elastography: A pilot study. J Ultrasound Med 2015;34:649–654. Chang JM, Moon WK, Cho N, Kim SJ. Breast mass evaluation: Factors influencing the quality of US elastography. Radiology 2011;259: 59–64. Chen XM, Cui LG, He P, Shen WW, Qian YJ, Wang JR. Shear wave elastographic characterization of normal and torn achilles tendons. J Ultrasound Med 2013;32:449–455. Chilappa CS, Aronow WS, Shapiro D, Sperber K, Patel U, Ash JY. Gout and hyperuricemia. Compr Ther 2010;36:3–13. Chowalloor PV, Keen HI. A systematic review of ultrasonography in gout and asymptomatic hyperuricaemia. Ann Rheum Dis 2013;72: 638–645. de Avila Fernandes E, Kubota ES, Sandim GB, Mitraud SA, Ferrari AJ, Fernandes AR. Ultrasound features of tophi in chronic tophaceous gout. Skeletal Radiol 2011;40:309–315. Dillman JR, Heider A, Bilhartz JL, Smith EA, Keshavarzi N, Rubin JM, Lopez MJ. Ultrasound shear wave speed measurements correlate with liver fibrosis in children. Pediatr Radiol 2015;45:1480–1488. Grgurevic I, Puljiz Z, Brnic D, Bokun T, Heinzl R, Lukic A, Luksic B, Kujundzic M, Brkljacic B. Liver and spleen stiffness and their ratio assessed by real-time two dimensional-shear wave elastography in patients with liver fibrosis and cirrhosis due to chronic viral hepatitis. Eur Radiol 2015;25:3214–3221. Leng QY, Tang YJ, Zhang LY, Xiang X, Qi L, Su BH. Diagnostic value of ultrasound imaging in chronic gouty arthritis. Sichuan Da Xue Bao Yi Xue Ban 2014;45:424–427. 446. Naredo E, Collado P, Cruz A, Palop MJ, Cabero F, Richi P, Carmona L, Crespo M. Longitudinal power Doppler ultrasonographic assessment of joint inflammatory activity in early rheumatoid arthritis: Predictive value in disease activity and radiologic progression. Arthritis Rheum 2007;57:116–124. Shi XQ, Li JL, Wan WB, Huang Y. A set of shear wave elastography quantitative parameters combined with ultrasound BI-RADS to assess benign and malignant breast lesions. Ultrasound Med Biol 2015;41:960–966. Smith HS, Bracken D, Smith JM. Gout: Current insights and future perspectives. J Pain 2011;12:1113–1129. Tausche AK, Jansen TL, Schr€oder HE, Bornstein SR, Aringer M, M€uller-Ladner U. Gout—Current diagnosis and treatment. Dtsch Arztebl Int 2009;106:549–555. Vlad M, Golu I, Bota S, Vlad A, Timar B, Timar R, Sporea I. Real-time shear wave elastography may predict autoimmune thyroid disease. Wien Klin Wochenschr 2015;127:330–336. Wang ZL, Li JL, Li M, Huang Y, Wan WB, Tang J. Study of quantitative elastography with supersonic shear imaging in the diagnosis of breast tumours. Radiol Med (Torino) 2013;118:583–590. Zhou J, Zhan W, Chang C, Zhang X, Jia Y, Dong Y, Zhou C, Sun J, Grant EG. Breast Lesions: Evaluation with shear wave elastography, with special emphasis on the ‘‘stiff rim’’ sign. Radiology 2014;272: 63–72.