Lucy Kerr, MD, Kerr Institute of Research and Teaching, Fabiano Mesquita Callegari, MD, Department of Pathology – Federal University of São Paulo , Deborah Rozenkwit, MD, Kerr Institute of Research and Teaching
Most breast cancer false negative (FN) by ultrasound examination (US) come from isoechoic tumor, along with lesions smaller than 5mm in diameter. We present the first US-FN that was avoided by a new imaging technique – Virtual Touch Elastography (VTE) with acoustic radiation force impulse (ARFI), that allowed detecting a small isoechoic breast cancer where microcalcifications were seen in mammography and also in US, but without tumor mass being identifiable.
A 58-year-old woman is followed in our department since November 1995, performing regular breast ultrasound. In family history, she has a 45 years old sister with breast invasive ductal carcinoma with axillaries metastasis. Menopausal since May 2001, without history of hormone therapy reposition. US exam from October 2010 was performed with 18 MHz probe showing many microcalcifications in the right upper outer quadrant (RUOQ) and no tumor mass was identified. However, mammography from September 2010 revealed clustered pleomorphic microcalcifications and architectural distortion in the RUOQ (Figure 1), without visible mass, classified as BIRADS V. Due to this result, an ultrasound review was requested (Figure 2) and was able to identify calcifications in the suspicious region but no tumoral mass. An elastographic study from RUOQ was conduct with VTE screening of each lobe (Figure 3) using ARFI technique, that has been incorporated into a conventional ultrassonographic devide (Acuson Siemens 2000; Siemens Medical Solutions, Mountain View, Calif), mainly where microcalcifications was US identified, and between 34mm and 46mm from the nipple, at 10 o’clock lobe, it was seen a dense well-defined mass in elastogram, with irregular and speculated borders. Since US showed no mass, the size and format of the lesion by US B-mode image were considered as the area containing calcifications at 10o´clock lobe, which measured 1.4×1.3×0.8cm in major axes. The tumor format in US B-mode image was quite different from the elastogram, which is considered a suspicious sign of malignancy by elastography 1.2. The lesion size was significantly larger on manual elastography (ME) and VTE than in US B-mode image. The largest diameter of malignant lesion of the right breast was 1.2cm, 2.1cm and 1.7cm; the mean diameter was 1.07cm, 1.80cm and 1.55cm and the mean area was 0.54cm², 1.47cm² and 1.82cm² by US B-mode image (considered as ACC), on US, ME and VTE, respectively. The shear-wave speed inside VTE mass was (Figure 4): 7.48 m/s, 7.50 m/s and 7.60 m/s (average speed = 7.52 m/s) and were significantly higher than in surround parenchyma nearby: 3.52m/s, 3.05 m/s and 2.73m/s (average speed = 3.10m/s), meaning the mass was stiffer than fibroglandular parenchyma. Size and velocity have also suspicious pattern of malignancy by elastography1,2 . The tumoral lesion, so easily identifiable by elastography, correlated to the area where suspicious microcalcifications was found, despite otherwise normal US B -mode image breast lobe appearance. The 2.42 times increase in shear-wave speed inside the lesion, when compared to normal surrounding parenchyma, indicates that it was harder and inelastic than glandular breast tissue, which is more common in malignant tumors. We concluded that all elastographic criteria were strongly suggestive of malignancy. Failure to identify the tumor in US B-mode image indicates that the lesion was isoechoic with normal breast parenchyma (ultrasound false-negative) and, at the same time, it was harder, allowing elastographic identification- an unusual combination. A fine needle aspiration biopsy (FNAB) was performed in september 23th, 10 guided by US and VTE, because the tumor was only identifiable by elastography. The cytological findings were consistent with the diagnosis of invasive ductal carcinoma grade 2. Some areas of cytology showed the classic pattern described for ductal carcinoma in situ, with rigid large ducts, cribriform pattern and microcalcifications, indicating that the lesion probably would have “in situ” and invasive areas. The background showed two cytological criteria of invasion, fibroelastoid fragments and fibroblastic proliferation among tumor cells, which ensures a positive predictive value for invasiveness of 96% ³. Surgery was done and an 11mm mass was found in the RUOQ – mainly in situ ductal carcinoma- with a small area of invasion posteriorly located.
The VTE-ARFI is a new tissue strain imaging technology that uses sound waves to interrogate the mechanical properties of tissue stiffness. VTE imaging is a qualitative gray scale map of relative tissue stiffness (elastogram). This information is computed by examining the relative displacement of tissue elements due to an acoustic pulse push. For a given elastogram image, bright regions depict tissue that is more elastic (less stiff) than dark regions. The VTE is displayed side-by-side with the corresponding conventional US B-mode image, but the tissue boundaries between the images may differ as they rely on different physical properties, which give different tissue contrast. The examination starts taking a baseline US which will give the anatomic orientation. Next, a push pulse is applied along this line to obtain the displaced tissue signal and VTE tissue quantification ARFI imaging involves targeting an anatomic region to be interrogated for elastic properties with use of a region-of-interest cursor while performing real-time B-mode imaging. In our case, as the region of interest could not be identify by US, it was done a screening in the entire RUOQ, each one was mechanically excited by using short-duration (∼200 μsec) acoustic pulses with a fixed transmit frequency of 5.00 MHz to generate localized tissue displacements. The displacements result in shear-wave propagation away from the region of excitation and are tracked by using US correlation-based methods 4. The maximal displacement is estimated for many ultrasound tracking beams that are laterally adjacent to the single push beam and transmitted at a nominal center frequency of 6.2 MHz and a pulse repetition frequency of 4–9 MHz. By measuring the time to peak displacement at each lateral location, one can reproduce the shear-wave speed of the tissue 5, 6. The shear velocity is estimated in a central window of 5mm axial by 4mm width within a region of interest graphically displayed at a size of 10mm axial by 6mm width. The shear-wave propagation speed is proportional to the square root of tissue elasticity 7,8. Results are expressed in meters per second. VTE-ARFI imaging was performed with a linear array at 9 MHz for US B-mode image and has demonstrated a lesion in the region of microcalcifications which is stiffer (less elastic) than the surrounding breast tissue, which was not seen in the US B-mode image. Because there were others microcalcifications foci in the breast tissue, only it was possible to identify the tumor after virtual elastography screening entire RUOQ. If the FNAB US-guided had been made without the aid of VTE, a false-negative (FN) citopathological study could have resulted if the sample come from an area of benign calcifications. It is well known that US cannot differentiate benign from malignant calcification and fibrosis foci can mimic microcalcifications. In most cases, US B-mode image suspects that a given calcification is malignant if a tumor mass, with several criteria of malignancy, is found around the calcifications. However, when the tumor is isoechoic with breast tissue, this differentiation is not possible. Most US FN come from isoechoic tumor, along with lesions smaller than 5mm9, 10,11. In this case, more important than describing the US difficulties to diagnosis breast cancer, is to show that they were overcome by the association US-VTE, which eliminated the US failure (FN). We also used ME to search the tumor (Figure 5), which did not offer the same clinical confidence than VTE, because it was slower, had many artifacts and the image continually modified, turning difficult to examine large regions and to obtain precise measurement of lesion (shape and dimensions of anatomic landmark changes with pressure and the boundaries are not clear). VTE imaging is more precise in capturing tissue strain map, which was obtained from a single push pulse, quick and quantifiable, reproducing the image in a similar manner every time, allowing better quality and image definition, given more diagnostic certainty. In this case, VTE used to investigate pathological calcification shown on mammogram and US, clearly demonstrated the underlying tumor that was not seen by US-B mode image, because the cancer was isoechoic (US- FN); on the other hand, in places where calcifications were benign, no tumor was observed on elastogram. To our knowledge, this case is the first and only one up to now, whose diagnosis of breast cancer was carried out by VTE screening (kind of pan-elastography), because US showed no tumor mass and could not explain mammography BIRADS V.
VTE is a non-invasive technique, without biological effects, that uses tissue stiffness information for diagnostic purpose, has easier acquisition of strain imaging, improves image quality and allows a new quantitative method that looks very promising to solve a serious US problem: its inability to identify isoechoic tumor where suspicious microcalcifications are found. VTE is more reproductible, easier to perform and faster than ME, allowing to scan large areas, as in this case. VTE could be used also to assist FNAB-US guidance – to avoid sample error. It is a new imaging procedure and more studies should be taken to corroborate our findings to determine if the association US B-mode – VTE is applicable on a large scale for screening breast cancer.