Evaluation of Virtual Touch Tissue Imaging Quantification (VTIQ - 2D-SWE) in the Assessment of BI-RADS® 3 and 4 Lesions

Evaluation of Virtual Touch Tissue Imaging Quantification (VTIQ - 2D-SWE) in the Assessment of BI-RADS® 3 and 4 Lesions

Evaluation of Virtual Touch Tissue Imaging Quantification (VTIQ - 2D-SWE) in the Assessment of BI-RADS® 3 and 4 Lesions: Can Patient Selection for Biopsy be Improved? - A Confirmatory Multi-Center-Study

The primary aim of this study is to evaluate if VTIQ in addition to BI-RADS® categorization can improve the diagnostic accuracy with respect to detection of malignancies, in particular for BI-RADS® categories 3 and 4a. The idea of the study is to restage all patients in categories 3 and 4a according to a predefined VTIQ cut-off value of ≥ 3.5 m/s (37 kPa).

Elastography is a method of imaging tissue stiffness. It is based on shear wave velocity information that can be mapped to create an image of the stiffness in the region of interest. Sonoelastography is used to differentiate benign from malignant lesions since malignant lesions alter tissue elasticity. Adding Shear Wave elastographic features to BI-RADS® feature analysis- especially in lesions scored BI-RADS® 3 and 4a- improved specificity of breast US mass assessment without loss of sensitivity. The BI-RADS® categories are defined by the risk for a malignant lesion varying from benign BI-RADS® 2 lesions, up to a 2% malignancy rate in BI-RADS® 3 and 2- 95% in BI-RADS® 4 (4a 2-10%; 4b 10-50%; 4c 50-95%). Based on these probabilities, biopsies are recommended for BI-RADS® 4 and 5 lesions and short-term follow-up examinations for BI-RADS® 3. Consequently, up to 2% of the in Ultrasound visible breast cancers are not directly detected as such and put into the BI-RADS® 3 category. In contrast, in the BI-RADS® 4a category more than 90% of the biopsies are unnecessary. The main aim of the confirmatory study is to use Virtual Touch Tissue Imaging Quantification in order to reduce unnecessary benign biopsies without a reduction of the number of detected cancers. This multi-center study is planned to involve 12 sites in 7 countries. Recruitment started at the first sites in February 2016. Recruitment takes place in the course of the patient's routine visit at a certified breast unit. All study participants will receive VTIQ in addition to standard ultrasound. Enrollment goal is a total of 1000 cases, split into groups of a minimum of n= 300 BI-RADS® 3, n= 400 BI-RADS® 4a, n= 100 BI-RADS® 4b, n= 100 BI-RADS® 4c. All patients will be documented in a screening list. Monitoring will be performed by the Coordination Center for Clinical Trials (KKS Heidelberg). Completeness, validity and plausibility of data will be checked in time of data entry (edit-checks) and using validating programs, which will generate queries. The investigator or the designated representatives are obliged to clarify or explain the queries. If no further corrections are to be made in the database it will be closed and used for statistical analysis. All data management procedures will be carried out on validated systems and according to the current Standard Operating Procedures (SOPs) of the Institute of Medical Biometry and Informatics. The standard BI-RADS® Ultrasound (US) category (BI-RADS® 3-4c) and VTIQ values will be correlated with the histological result. Additionally, local (BI-RADS® given at each site) and central expert BI-RADS® assessment will be compared (BI-RADS® assessment and assessment of the variables leading to the BI-RADS® value separately) to assess the inter-rater reliability. In addition, the BI-RADS® assessments will be compared with the histological results. The variable "measurement lesion (in m/s)" is derived from three VTIQ measurements as follows: I. For confirmatory analysis of primary objectives an algorithm was established, using the first measurement for analysis if the value is smaller than 2.5 m/s or if the value is larger than 4.5 m/s. If the first measurement is smaller than 2.5 m/s, the lesion can be considered benign and no further diagnostics is needed. If the lesion is larger than 4.5 m/s the lesion should be considered suspicious and further diagnostics is required. requiring two additional measurements (in total three measurements) if the first measurement is in the range of ≥ 2.5 m/s to ≤ 4.5 m/s. In this case the average of all three measurements is used for analysis. II. For descriptive analysis other options for derivation of this variable from the three VTIQ measurements will be calculated for discussion: First measurement only Average of all three measurements Median of all three measurements Maximum of all three measurements In conjunction with the maximum VTIQ shear wave velocity the quality display will be used to aid in the classification of lesions as malignant or benign as follows: If the shear wave velocity ≥ the cut-off value (3.5 m/s), the lesion is considered potentially malignant, regardless of the outcome of the quality factor If a lesion or lesion rim has a completely high quality factor (all green) and a shear wave velocity < the cut-off value (3.5 m/s), the lesion is considered benign. If a lesion or lesion rim has mixed high and low quality factors (there are areas with low quality within the mass) and the only area of high quality (green) has a shear wave velocity < the cut-off value (3.5 m/s), then the lesion is indeterminate and malignancy cannot be excluded.

Virtual Touch Tissue Imaging, VTIQ, Elastography, Breast Cancer, BI-RADS, Differentiation of breast lesions, Breast Lesion

Siemens Medical Solutions USA, Inc. (Mountain View, CA) has implemented Virtual Touch Tissue Imaging Quantification (VTIQ) technology on a commercially available general purpose US imaging system (trade name: Acuson S2000 or S3000). This system has received clearance under Food and Drug Administration (FDA) 510(k) number K072786 (S3000) and K130881 (VTIQ). The technology uses a set of tailored US pulses (Acoustic Radiation Force Impulse, ARFI) to induce shear waves in breast tissue due to tissue displacement. A set of standard B-mode pulses detect the perpendicular shear waves. The displacement signals can be processed using algorithms on a Virtual Touch IQ-equipped system in order to calculate the shear wave velocity.

Standard chi-square test at a two-sided significance level of 5% to test the diagnostic accuracy of Virtual Touch Tissue Imaging Quantification (VTIQ) in the differentiation of BI-RADS® 3 and 4a lesions

Hierarchical testing of the main statistical hypotheses H0_1, H0_2, H0_3 (test for H0_1 defines a gatekeeper) to evaluate the diagnostic accuracy of VTIQ in the differentiation of BI-RADS® 3 and 4a lesions: H0_1: The proportion of malignancies for group A (BI-RADS 4a, VTIQ ≥ cut-off) is less or equal to the proportion of malignancies for group B (BI-RADS 4a, VTIQ < cut-off). versus H1_1: The proportion of malignancies for group A is higher than the proportion of malignancies for group B.

H0_2: The proportion of malignancies for group B (BI-RADS 4a, VTIQ < cut-off) is larger than or equal to 2%. versus H1_2: The proportion of malignancies for group B is smaller than 2%.

H0_3: The proportion of malignancies for group D (BI-RADS 3, VTIQ < cut-off) is larger than or equal to 2%. versus H1_3: The proportion of malignancies for group D is smaller than 2%.

Corresponding chi-square tests for descriptive analysis of Virtual Touch Tissue Imaging Quantification (VTIQ) in the assessment of BI-RADS® 3, 4b, 4c lesions

Equivalent hypotheses as given for primary outcome measures can be formulated for the patients finally staged as BI-RADS® 3, 4b and 4c. These hypotheses will be analyzed with corresponding chi-square tests, where the resulting p-values are only interpreted descriptively: To assess whether ultrasonically visualized breast lesions, categorized as BI-RADS® 3/ 4b/ 4c with a VTIQ-measured shear velocity value of smaller than 3.5 m/s (37 kPa), show a lower malignancy rate than BI-RADS® 3/ 4b/ 4c with a VTIQ-measured shear velocity value of larger than or equal to 3.5 m/s (37 kPa).

Two logistic regression models including including BI-RADS® or BI-RADS® and VTIQ will be compared to test whether the probabilities for malignancies predicted with ultrasonically visualized breast lesions categorized as BI-RADS® differ from the probabilities for malignancies predicted with ultrasonically visualized breast lesions categorized as BI-RADS® and VTIQ-measured shear velocity.

Corresponding chi-square tests for descriptive analysis of the strain ratio in the differentiation of BI-RADS® 3, 4a, 4b or 4c lesions

Equivalent hypotheses as given in H0_1 vs H1_1 can be formulated for the patients initially staged as BI-RADS® 3, 4b and 4c. These hypotheses will be analyzed with corresponding chi-square tests, where the resulting p-values are only interpreted descriptively to test whether for women with ultrasonically visualized breast lesions categorized as BI-RADS® 3, 4a, 4b or 4c respectively, the subgroup of women with a strain ratio smaller than or equal to 1 shows a lower malignancy rate than the subgroup of women with a strain ratio of larger than 1.

Robust regression models (Passing-Bablok-regression, orthogonal regression) and Bland-Altman Plots will be used for pairwise comparison of the continuous values between raters, where a slope of 1 and an intercept of 0 indicates agreement.

Robust regression models (Passing-Bablok-regression, orthogonal regression) and Bland-Altman Plots will be used for pairwise comparison of the continuous values between raters, where a slope of 1 and an intercept of 0 indicates agreement.

Thereby the following factors will be examined: Subject-related factors: Skin - breast lesion surface depth (cm) Quality factor (color coded scale) within the lesion Breast density/ Tissue composition (homogeneous background texture fat, homogeneous background texture fibroglandular, heterogeneous background texture) Lesion size in B-mode (cm) Normal fatty tissue shear wave velocity (Ratio between measurement in the fatty tissue and in the lesion) Pathology (fibroadenoma, lipoma, atypia, cyst condense, "non special type" (NST), invasive lobular carcinoma (ILC), invasive tubular carcinoma (ITC), carcinoma with medullary features, papillary cancer, ductal carcinoma in situ (DCIS), others) Grading (G1, G2, G3, Gx) Immunohistology (estrogen receptor (ER) status positive/ negative, progesterone receptor (PgR) status positive/ negative, human epidermal growth factor receptor 2 (HER2-neu) status positive/ negative, Ki-67 status (%))

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the skin - breast lesion surface depth (cm).

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the Quality factor (color coded scale) within the lesion.

Breast density/ tissue composition assessed using the morphologic characteristics on US (homogeneous background texture fat, homogeneous background texture fibroglandular, heterogeneous background texture)

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the breast density/ tissue composition (homogeneous background texture fat, homogeneous background texture fibroglandular, heterogeneous background texture).

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the lesion size in B-mode (cm).

Normal fatty tissue shear wave velocity (ratio between measurement in the fatty tissue and in the lesion)

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the normal fatty tissue shear wave velocity (ratio between measurement in the fatty tissue and in the lesion).

Pathology (fibroadenoma, lipoma, atypia, cyst condense, "non special type" (NST), invasive lobular carcinoma (ILC), invasive tubular carcinoma (ITC), carcinoma with medullary features, papillary cancer, ductal carcinoma in situ (DCIS), others)

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the pathology (fibroadenoma, lipoma, atypia, cyst condense, "non special type" (NST), invasive lobular carcinoma (ILC), invasive tubular carcinoma (ITC), carcinoma with medullary features, papillary cancer, ductal carcinoma in situ (DCIS), others).

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the grading (G1, G2, G3, Gx).

Immunohistology (estrogen receptor (ER) status positive/ negative, progesterone receptor (PgR) status positive/ negative, human epidermal growth factor receptor 2 (HER2-neu) status positive/ negative, Ki-67 status (%))

For descriptive analysis of predictive factors of the continuous VTIQ-value, different univariable und multivariable linear regression models will be evaluated and compared. Thereby one factor to be examined is the immunohistology (estrogen receptor (ER) status positive/ negative, progesterone receptor (PgR) status positive/ negative, human epidermal growth factor receptor 2 (HER2-neu) status positive/ negative, Ki-67 status (%)).

Cohens's Kappa to determine the inter-rater reliability of BI-RADS® Assessment (local vs. central assessment)

Local (BI-RADS® given at each site) and central expert BI-RADS® assessment will be compared (BI-RADS® assessment and assessment of the variables leading to the BI-RADS® value separately). To test the inter-rater reliability of BI-RADS® Assessment Cohens' Kappa will be calculated.

Interclass correlation coefficient (ICC) values to determine the inter-rater reliability of BI-RADS® Assessment (local vs. central assessment)

Local (BI-RADS® given at each site) and central expert BI-RADS® assessment will be compared (BI-RADS® assessment and assessment of the variables leading to the BI-RADS® value separately). To test the inter-rater reliability of BI-RADS® Assessment ICC values will be calculated.

To compare the BI-RADS® assessments with the histological results, the malignancy rates will be computed separately for each BI-RADS® value. It will then be checked whether these rates lie within the given intervals. Furthermore, the likelihood of malignancy estimated by the experts will be examined using receiver operating curve (ROC) and area under the curve (AUC).

Chi-square test to examine whether the cut-off value of ≥ 3.5 m/s (37kPa) might be increased by increasing the cut-off value step-by-step by a small amount and repeat the primary analysis

We will increase the cut-off value step-by-step by a small amount and repeat the primary analysis (using a chi-square test). We stop the testing procedure as soon as the p-value is above the significance level (0.05). This analysis is, obviously, only an explorative one and has no confirmatory value.

Inclusion Criteria: Female Age ≥18 years Patients with a lesion ≥ 0.5 cm in largest diameter size, initially scored BI-RADS® 3, 4a, 4b or 4c in B-mode ultrasound Informed consent about histological examination (core cut biopsy (CCB), vacuum-assisted biopsy (VAB), fine needle aspiration (FNA) or surgery) has already been given in the course of clinical routine Signed informed consent of study participation Exclusion Criteria: Pregnant or lactating women Women with breast implants on the same side as the lesion Women that underwent local radiation or chemotherapy within the last 12 months Women with history of breast cancer or breast surgery in the same quadrant Lesions in or close to scar tissue (< 1cm) Skin lesions or lesions that have been biopsied previously Lesion larger than 4 cm in the longest dimension No lesion should be included when more than 50% of the lesion is further down than 4 cm beneath the skin level.

Universitätsklinikum Marburg, Klinik für Gynäkologie, gyn. Endokrinologie und Onkologie Senologische Diagnostik & Gynäkologischer Ultraschall

Barr RG, Zhang Z. Shear-wave elastography of the breast: value of a quality measure and comparison with strain elastography. Radiology. 2015 Apr;275(1):45-53. doi: 10.1148/radiol.14132404. Epub 2014 Nov 24.

Barr RG, Zhang Z. Effects of precompression on elasticity imaging of the breast: development of a clinically useful semiquantitative method of precompression assessment. J Ultrasound Med. 2012 Jun;31(6):895-902. doi: 10.7863/jum.2012.31.6.895.

Golatta M, Schweitzer-Martin M, Harcos A, Schott S, Junkermann H, Rauch G, Sohn C, Heil J. Normal breast tissue stiffness measured by a new ultrasound technique: virtual touch tissue imaging quantification (VTIQ). Eur J Radiol. 2013 Nov;82(11):e676-9. doi: 10.1016/j.ejrad.2013.06.029. Epub 2013 Aug 8.

Golatta M, Schweitzer-Martin M, Harcos A, Schott S, Gomez C, Stieber A, Rauch G, Domschke C, Rom J, Schutz F, Sohn C, Heil J. Evaluation of virtual touch tissue imaging quantification, a new shear wave velocity imaging method, for breast lesion assessment by ultrasound. Biomed Res Int. 2014;2014:960262. doi: 10.1155/2014/960262. Epub 2014 Mar 31.

Tozaki M, Saito M, Benson J, Fan L, Isobe S. Shear wave velocity measurements for differential diagnosis of solid breast masses: a comparison between virtual touch quantification and virtual touch IQ. Ultrasound Med Biol. 2013 Dec;39(12):2233-45. doi: 10.1016/j.ultrasmedbio.2013.07.012. Epub 2013 Sep 21.

Barr RG, Nakashima K, Amy D, Cosgrove D, Farrokh A, Schafer F, Bamber JC, Castera L, Choi BI, Chou YH, Dietrich CF, Ding H, Ferraioli G, Filice C, Friedrich-Rust M, Hall TJ, Nightingale KR, Palmeri ML, Shiina T, Suzuki S, Sporea I, Wilson S, Kudo M. WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 2: breast. Ultrasound Med Biol. 2015 May;41(5):1148-60. doi: 10.1016/j.ultrasmedbio.2015.03.008. Epub 2015 Mar 18.

Cosgrove DO, Berg WA, Dore CJ, Skyba DM, Henry JP, Gay J, Cohen-Bacrie C; BE1 Study Group. Shear wave elastography for breast masses is highly reproducible. Eur Radiol. 2012 May;22(5):1023-32. doi: 10.1007/s00330-011-2340-y. Epub 2011 Dec 31.

Balleyguier C, Canale S, Ben Hassen W, Vielh P, Bayou EH, Mathieu MC, Uzan C, Bourgier C, Dromain C. Breast elasticity: principles, technique, results: an update and overview of commercially available software. Eur J Radiol. 2013 Mar;82(3):427-34. doi: 10.1016/j.ejrad.2012.03.001. Epub 2012 Mar 24.

Barr RG, Destounis S, Lackey LB 2nd, Svensson WE, Balleyguier C, Smith C. Evaluation of breast lesions using sonographic elasticity imaging: a multicenter trial. J Ultrasound Med. 2012 Feb;31(2):281-7. doi: 10.7863/jum.2012.31.2.281.

Barr RG. Sonographic breast elastography: a primer. J Ultrasound Med. 2012 May;31(5):773-83. doi: 10.7863/jum.2012.31.5.773.

Evans A, Whelehan P, Thomson K, Brauer K, Jordan L, Purdie C, McLean D, Baker L, Vinnicombe S, Thompson A. Differentiating benign from malignant solid breast masses: value of shear wave elastography according to lesion stiffness combined with greyscale ultrasound according to BI-RADS classification. Br J Cancer. 2012 Jul 10;107(2):224-9. doi: 10.1038/bjc.2012.253. Epub 2012 Jun 12.

Golatta M, Pfob A, Busch C, Bruckner T, Alwafai Z, Balleyguier C, Clevert DA, Duda V, Goncalo M, Gruber I, Hahn M, Kapetas P, Ohlinger R, Rutten M, Togawa R, Tozaki M, Wojcinski S, Rauch G, Heil J, Barr RG. The potential of combined shear wave and strain elastography to reduce unnecessary biopsies in breast cancer diagnostics - An international, multicentre trial. Eur J Cancer. 2022 Jan;161:1-9. doi: 10.1016/j.ejca.2021.11.005. Epub 2021 Dec 5.