Tổng số lượt xem trang

Thứ Sáu, 5 tháng 2, 2016

MEDIC ARFI in BREAST TUMORS

CLINICAL FINDINGS of ARFI in BREAST TUMORS

VO NGUYEN THUC QUYEN, PHAN THANH HAI, MEDIC MEDICAL CENTER,
HCMC, VIETNAM
INTRODUCTION:
Breast Cancer is currently the top cancer among women worldwide including Viet nam. Therefore, early detection plays a critical role in clinical decision of management.
Besides Mammography and MRI, ultrasound has been a useful modality in detecting breast tumors. Moreover, the combination with Color Doppler significantly reinforces the B-mode diagnosis. Lately, new ultrasound technique, elastography is providing more information to increase accuracy. However, each one uses different method including compressed and non-compressed technologies. Developing by Siemen, ARFI is a non-compressed elastography, evaluates tissue stiffness base on replacement caused by acoustic radiation force impulse (ARFI). In other words, tissue deformed and reformed  under a force. The stifferness replaces less compared with surrounding tissue in same depth. In clinical application, tumors usually harder than healthy tissue.
AIMS:
To evaluate ARFI qualitative and quantitative assessment to differentiate benign and malignant breast tumors.
METHODS and MATERIALS
Patient and Pathologic diagnosis:
From April to November 2015, we selected 85 breast lesions classified as category 3-5 according to ACR Breast Imaging Recording and Data System (BI-RADS). Two radiologists analyzed them in the following steps before performed biopsy with final diagnosis (FNAC, Core Biopsy, Excisional Biopsy). All images and biopsy procedures were performed at Medic Medical Center Ho Chi Minh city. Exclusion criteria include:
·         Non histopathology confirmation
·         Male breast lesions
Imaging methods:
Using linear probe 9L4 (9MHz) in Siemens Acuson S2000, we applied respectively 2 modes:
·         VTI (Virtual Touch Quantification): an gray-scale elasticity map within region of interest (ROI)
·         VTQ: (Virtual Touch Quantification): quantitatively measure shear-wave speed (m/s) within non-resizable ROI. The ROI was set in multiple point of the lesion to get the mean measurement.
Step 1: scan B-mode and Color Doppler images, classified lesion using BI-RADS lexicon (shape, orientation, border, echotexture, posterior feature)
 Step 2: Acquired Elasticity Score (E.S) in VTI mode then measure Area Ratio (proportion between VTI lesion area and B-mode area). Base on VTI map, we classified lesions with 5 elasticity score: Figure
            Score 1: totally white
            Score 2: mosaic (mix multi-shade of grey and white)
            Score 3: black core with white or grey or mix
            Score 4: totally or near to complete black
            Score 5: totally black with black component out of lesion
Score1-3: low suspect of malignancy
Score 4-5: high suspect of malignancy

Step 3: Set ROI in 5 different points of the lesion then measured Shear-wave Velocity (SWV) in VTQ mode. We calculated mean velocity for each lesion. The ROI in VTQ mode are fixed with 5 x 5 mm in size. When acquired velocity reach over 9.10m/s or computer is unable to get the signal, we have X.XX m/s as value. [1] Figure 2.







Figure 2: Shearwave travels through hard tissue very fast with > 9.10m/s (X.XX m/s value)
Statistic analysis:
We use SPSS version 16.0 to identified cut-off value and obtain ROC for best value of sensitivity and specificity. Once we get cut-off value, we use t-student analysis to see whether benign and malignant populations were statistically different.
           
RESULT
This study was approved by the institutional review board and informed consent was obtained from all participants. From April to November 2015, we selected 85 breast lesions including 59 benign (69.4%) and 26 (30.6%) malignant. Lesions appear to dominantly locate in right breast 52/85 (61.2%), left 33/85 (38.8%). The mean size 16.26 ±6.56 width and  9.64 ±5.01 mm depth
Histopathologic diagnosis
n (%)
Malignant:  Invasive ductal carcinoma
26 (29.4)




Benign
59 (70.6)
Fibroadenoma
5 (5.9)
Mastitis
3 (3.5)
Intraductal papilloma
2 (2.4)
Fibrocystic change
46 (55.3)
Others
3 (3.5)
Total
85 (100)
Table 1: histopathlogic diagnosis of malignant and benign breast lesions
ARFI analysis
-VTI:
a/ Elasticity Score (E.S)

Malignant (%)
Benign (%)
ES 1
0
0
ES 2
0
52.5
ES 3
0
47.5
ES 4
23.1
0
ES 5
76.9
0
Total
100
100
Table 2.1: ES Score frequency of malignant and benign lesion
As the table 2.1, 26/26 cancer cases has ES 4-5 within suspicious range.
b/ Area ratio (A.R)
Area Ratio
Sensitivity (%)
Specificity (%)
1.06
100
27.2
1.13
96.2
52.5
1.20
88.5
64.4
1.34
88.5
94.9
1.40
84.6
96.6
1.44
84.6
98.3
Table 2.2:

As the table 2.2, the AR cut-off point would best at 1.34 with sensitivity 88.5% and specificity 94.9%. Area under ROC curve for malignancy is 0.933.
-VTQ:
We excluded 8 malignant cases has SWV as X.XX m/s
SWV
Sensitivity (%)
Specificity (%)
2.20
100
69.5
2.24
94.4
72.9
2.32
88.9
79.7
2.41
83.3
83.1
2.49
77.8
88.1
Table 2.3:
As the table 2.3, the SWV cut-off point would best at 2.24 with sensitivity 94.4% and specificity 72.9%. Area under ROC curve for malignancy is 0.911.
DISCUSSION
The ability of early detection
ARFI helps in differentiate malignant and benign lesion. E.S score in VTI mode suggest suspicion are quite accurate in this study (26/26). The gray-scale map not only distinguish big tumors but also in small tumors as case demonstrated (Figure 3). It could greatly aid in early detection.
Figure 3: A DCIS 6 x 5 mm mass with BI-RADS 5 in B-mode and ES 5, infiltration is clearly demonstrated which is not visible on conventional B-mode.
In term of quantitative evaluation, Area Ratio reinforced E.S. It also shows a better the cancerous infiltration in surrounding tissue than conventional method. In conventional ultrasound, only when halo rings, architecture distortion, skin changes suggest infiltration. However, those present in late stage while we are aiming for early detection. (Figure 4)
Firgure 4: non-halo tumors with AR=1.81 is better demonstrated the surrounding invasion
Our cut-off value
Our SWV cut-off point at 2.24 m/are suitable for clinical practice. Other reference studies were significantly higher (Yoon Seok Kim et al: 4.23±1.09 m/sec [2]) as they considered all X.XX value as 9.10m/s. We excluded all X.XX value since it not actually equals 9.10m/s.
Role in clinical diagnosis
In clinical application, ARFI increases the accuracy of B-mode and Color Doppler. It most value in BI-RADS 3-4a lesion which are the borderline between benignity and malignancy. We recommended grade up from BI-RADS 3 to 4A if all ARFI features are suspicious. However, here are some exceptions. Acknowledged that some cancer such as Inflammatory Breast Cancer (IBC) tends to be softer than normal tissue, reversely, some benign condition like Mastitis can mask malignancy (figure 5). Our study limited in 85 case and not included any IBC however caution should be made if specially AR> 1.34. An interesting study was held by M.Teke et al. which used ARFI to compare Idiopathic Granulomatous Mastitis with Breast Cancer . Study shown significantly different between their SWV (cut-off value 4.08m/s with 80.6% sensitivity, 86.4% specificity). It is important not to miss cancer but still minimalize invasive option. EFSUMB also recommend this concept but less certain in down grade. In some situation, we can down grade 4A lesion if the technique done right, such as circumscribed lesion with suspicious Doppler pattern or posterior feature. ARFI also helps guiding FNA procedure as we puncture the hardest points in the lesion on VTI map.  

Figure 5: Mastitis lesion in 60 years old patient, BI-RADS 4C E.S 2, AR=1.1 and VTQ=1.58m/s


Technical recommendation


CONCLUSION
Overall, ARFI is a useful tools for diagnosis and biopsy guidance breast tumors. The technique is simple since it is non-compressed and repeatable. It cannot replaced biopsy but reinforced conventional ultrasound. This is a promising technique helps avoiding invasive diagnosis if we use it right and well-combined with other features.

REFERENCES:
1/ Wojcinski S, Brandhorst K, Sadigh G, Hillemanns P, Degenhardt F. Acoustic radiation force impulse imaging with Virtual TouchTM tissue quantification: mean shear wave velocity of malignant and benign breast masses. International Journal of Women’s Health. 2013;5:619-627. doi:10.2147/IJWH.S50953.
2/ Kim YS, Park JG, Kim BS, Lee CH, Ryu DW. Diagnostic Value of Elastography Using Acoustic Radiation Force Impulse Imaging and Strain Ratio for Breast Tumors. Journal of Breast Cancer. 2014;17(1):76-82. doi:10.4048/jbc.2014.17.1.76.
3/ M. Teke, M. Gümüş, F. Teke. Combination of elastography and tissue quantification using the acoustic radiation force impulse technology for differential diagnosis of Idiopathic Granulomatous Mastitis with Breast Cancer. ECR 2015 http://dx.doi.org/10.1594/ecr2015/C-1835

4/D. Cosgrove1, F. Piscaglia2, J. Bamber3. EFSUMB Guidelines and Recommendations on the Clinical Use of Ultrasound Elastography.Part 2: Clinical Applications. Ultraschall in Med 2013 DOI: 10.1055/s-0033-1335375
Đăng nhận xét