European Journal of Cardiovascular Medicine

India has the one third of the global breast cancer burden with China and the United states. The incidence of breast cancer is increasing in India. Most of patients are present in the advanced stage and thus making the treatment difficult. The mortality rate due to breast cancers is high in India as compared to developed and other developing countries. Both the locally aggressive lesions and distant metastases are the major causes of death in breast cancer¹.Clinical palpitation is used to determine the stiffness of the lesion since ancient Greeks and Egyptians². The stiff, non-mobile lesions of breast have chances of being malignant.  However in vitro experiments have shown that the difference in stiffness of malignant breast lesions and benign breast lesion is important but there is little overlap³. Thus elastography has high accuracy of characterization if breast lesions.

Lately a pathologic examination is the standard investigation to differentiate between benign and malignant breast mass. But it is an invasive procedure and it yields a benign diagnosis (close to 75%) in majority of the patients, thus it is challenging to deploy it in cancer screenings¹.

Elastography can improve the specificity of breast sonography by differentiating benign and malignant masses thus reducing the numbers of benign biopsies. The comparison between the elastography parameters like modified color score and strain ratio was done using trichromatic blue green and red (BGR) artifacts. The theory that works here is it images the gradient of the tissue displacement that happens following compression. Thus to differentiate between the benign and malignant masses a new modified version of color score was recommended through this research along with the mean strain score of 3:2.

Types of elastography

There are two types of elastography, ultrasonic strain elastography (SE) and shear wave elastography (SWE)⁴. Strain elastography is a qualitative technique and it evaluates the changes in tissues when an external force is applied. Softer tissues get deformed more easily than the stiffer tissues. There are two methods of applying stress, one is by manually compressing the tissue followed by release, second is by using an acoustic radiation force impulse (ARFI) push pulse.

Strain elastography is now used to assess various structures like breast, prostrate, liver, blood vessels, thyroid and musculoskeletal structures. The breast is the most widely and successfully examined organ using elastography as the imaging modality. Breast cancers are usually harder than normal breast tissue. Elastography can provide information about the tissue stiffness⁵. In strain elastography the relative stiffness in the tissue in response to the force applied over it is calculated. The systems vary on the basis of degree of stress that is required for optimal images. In some systems minimal compression and release is required while others require a moderate amount of compression and release. There is spot where the algorithm is the most accurate. Too much or too little compressions lead to the elastogramimages that are sub-optimal⁶.

It was a single center prospective observational study done on 80 cases and 100 palpable breast lumps. All the patents that had clinically palpable lump were included in the study. The patients that had prior biopsy proven recurrent breast cancers were excluded from the study. The study was done by using the ultrasound equipment with a 5 to 12 MHz linear array transducer by qualified radiologists.The cases were observed using B mode ultrasonography and followed by elastography subsequently. The ultrasonography was performed in supine position in radial fashion. The morphology of breasts observed under ultrasonography were interpreted on the basis of American College of Radiology breast Imaging Reporting and Data System (ACR BI-RADS 5^th Edition). In elastography the transducer was placed perpendicular to the lesion and it was compressed in a gentle manner. A modified color score and the strain ratio was used to determine the nature of the lesion. The modified color score uses a color pattern to determine the nature of the lesion, wherein the blue indicates malignancy i.e. hard lesions, red indicating the benign lesions which are the soft lesions and the green indicating the strain which is similar the normal adjacent tissue. The blue color was given a score of 2, uniform green color with BGR artifacts were scored of 0 and the mix if blue and green was given a score of 1.

Strain ratio is the ratio of the hardness of tissue of the lesion as compared to the normal adjacent breast fat. Lesions with higher score were considered as malignant while lower strain ratios are considered as benign. Strain ratio was calculated by placing the placing the first region of interest i.e. ROI 1 on the normal breast fat and the similar sized ROI 2 was placed over the lesion. Both the regions of interest should be placed horizontally. The values generated from the both the ROI’s were used to calculate the strain ratio. The cut off strain ratio used to differentiate between benign and malignant lesions was 3.3. The cut off value was calculated from the initial work done in 6 months and had shown a good correlation for differentiating benign and malignant lesions. The consent was taken from the cases when the pathology was observed on the B-mode sonogram and elastography images that were included in the study. Ultrasound guided fine needle aspiration (FNA) or True cut biopsy was done in the pathology department by the qualified pathologist. The results of the FNAC and biopsy were compared with the ultrasound and elastographic findings.

Statistical Analysis

Microsoft excel was used to enter the data. The measurable values were analyzed and interpreted between them by the student’s t test while the categorical variables between them were interpreted by Chi- square test (χ2) test. The receiver operating characteristic (ROC) curves for mean strain ratio and cut off values were obtained. The statistical procedures were performed using MedCalcver 19.1.7 and OpenEpiver 3.01. P values less than 0.05 were considered as statistically significant.

The study sample had 80 cases with 100 breast lesions, amongst the 80 cases there were 79 female one male. The mean age group was 41.67 +/- 15.8 years and the age ranged from 14 to 95 years. When BIRADS was considered the average age increased with the increasing BIRAD score and was statistically significant. A similar association was seen and it was statistically significant as well, was of modified color score and mean strain ratio with advancing age, Table 1. In this study,the benign lesions constituted for 52% of cases (n=52), while remaining 48.75% cases where n=48 cases were of malignant cases. Amongst the 52 benign cases, 32 were benign solid lesions and the remaining 20 cases were benign cystic lesions. In the age group of 41 and 60, there was an advancing trend in malignancy as there were higher number of lesions in the BIRADS category 4 and 5 which accounted for 72.9% of lesions with a modified color score of 2and the strain ratio of more than 3.2. The pathologic interpretation had shown that there were 27 malignant lesions. The people with age group more than 60 years there were 9 lesions and out of which 7 i.e. 77.78 % were classified as malignant on ultrasound (BIRADS 4 and 5) and had a modified color score of 2 was seen. On the basis of mean strain ratio all the case classified as malignant, were also confirmed by the pathologic diagnosis, while only 3 lesions were classified as benign.

The cystic benign lesions had shown the classic BGR artifacts on the modified color scoring scale in elastography. Amongst the 19 cystic benign lesions, 9 lesions had a final diagnosis of fibrocystic disease and the other 9 lesions were diagnosed as the breast abscess. The pathologic investigation used for final diagnosis were either FNA cytology or histopathologic examination. In 100 breast masses where either the cytology or histopathology were performed, there were 47 malignant and 53 benign breast masses, Table 2. In 100 breast masses, there were 48 cases that had the modified color score of 2. In these 47 cases, 46 were pathologically proven malignant cases and 1 were benign masses. There were 29 breast cases that had a modified color score of 1, there were 28 benign lesion and there was only one malignant lesion. While in 23 cases with modified color score as 0, all the cases were diagnosed as benign.

The modified color score had shown the sensitivity of 63.67%, specificity of 74.23%, while positive and negative predictive values being 85% and 45.9% respectively to differentiate benign from malignant lesions with overall diagnostic accuracy up to 67.45%. Through researches the modified color score has been adjusted and new modified dual color scoring system was formed. In new modified color score there only two categories benign and malignant. The Table no 3 shows the results obtained from the new modified dual color scoring system where the modified color scores of 0 and 1 (Tsukuba scores 1 to 3) were considered as benign while the modified color score of 2 (Tsukuba scores 4 and 5) were considered as malignant. This modification gave the sensitivity of 96.4%, specificity of 85.78%, positive predictive value of 85.92% and negative predictive value of 86.87 % and a diagnostic accuracy of 91.67%. The overall diagnostic accuracy has been improved as compared to three category system. Only case that was categorized as benign and was scored zero but on pathologic confirmation it was shown as malignant. Thus with new modified color scoring system the chance of missing a malignant lesion was 2.2%.

A cutoff strain ratio of 3.2 was used to differentiate the benign and malignant lesions. Out of 100 breast lesions, 55 breast masses had a strain ratio of 3.2 which indicates that the masses were malignant as shown in the Table 1. Mean strain ratio for solid benign lesions was 1.828 +/-1.187 (mean +/-SD), for cystic benign lesions it was 1.208 +/-0.786 (mean +/-SD) while for malignant masses it was 4.567+/- 0.928 (mean +/-SD). Thus overall mean strain ratio for benign lesions 1.678 +/- 1.123 which was statistically significant as compared to cut off strain ratio for malignant lesions (P<0.001).

It noted that the mean strain ratio of more than 3.2 has the sensitivity of 100% (95 % CI, 91.3%– 100%) and specificity of 97.91% (95% CI, 88.6% - 100%) to differentiate benign from malignant lesions. This is so significant that with the mean strain ratio of more than 3.2 none of the malignant lesions were missed.Mean strain ratio had identified all 47 malignant masses in correct manner. All the malignant cases were categorized as malignant, only one benign case was categorized as malignant. All the cases that were classified as benign were diagnosed as benign via histopathology. Pathological examination confirmed the 34 solid benign lesions and 19 cystic benign lesions. Ultrasonography was accurate in picking only 32 solid benign lesions and 16 cystic benign lesions, whereas elastography diagnosed 33 solid and 18 cystic masses correctly. Table no 2 has shown the comparison of B-mode ultrasonography and elastography with the pathologic correlation. The mean strain ratio was used for the final comparison between sonography and elastography and it also an objective measurement. All the 47 malignant lesions were diagnosed accurately via elastography while ultrasound was able to diagnose only 43 malignant masses. Thus the sonography had sensitivity of 88.7 %, specificity of 95.28%, PPV of 95.37 % and NPV of 89.18% and overall diagnostic accuracy of 92.89 %.

To diagnose a breast lesion as benign or malignant is very important clinically to avoid unnecessary biopsies. Elastogtraphy technique helps to differentiate between benign and malignant breast masses. It role is very important in the lesions with BIRADS score 3 and 4.  Elastography also reduces the unwanted biopsies particularly in benign and /or cystic lesions.

The most important age group in the study was 41 to 60 years, followed by 21-40 years with a mean age of 41.67 +/- 15.8 years (mean +/- SD) with a range of 14 to 95 years. The results were similar to the study by Menezes et al who said that the most common age of the women with breast lesions is 45 to 60 years of age⁷.  A similar observation was given by Dong et al. They reported the mean age being 45.39 +/-12.97 years among women with breast lesions⁸.

In the current study the most of the breast lesions are categorized under BIRADS 4 where n=32, which constitute 40 % percent of the cases. YilmazE et al, also showed that BIRADS 4 lesions constituted majority of the cases i.e. 43.09% of the total cases⁹. In the current study 53 i.e. 53% of the masses are benign and 47 i.e. 47 % of the total cases are malignant.Ozel and Ozel in their study reported that there is high incidence of benign lesions i.e. 71% of which benign solid lesions constituted 47.8 % of overall lesions while 23.3% were of cystic benign lesions. Malignant lesions constituted only 28.9% of overall cases¹⁰. The current findings also differ from observations made by Xiao et al, who observed that cystic lesions were of 39% of overall cases and solid benign lesions formed about 25.3 % of overall lesions. They also observed that there were more benign lesions i.e64.2 %¹¹

In the current study all the cystic lesions had shown the BGR artefact. Cho et al, in their study had seen 13 cystic lesions had shown BGR artefacts with strain elastography. The author also states that presence of trichromatic pattern or BGR artefacts helps in differentiating a cystic lesion from solid mass.

A modified color score was used as one of the parameters in the study. It is the modified version of Tsukuba scoring. Chang et al I their study had proposed the modification of Tsukuba score into categories as negative (score 0), borderline (score 1) and malignant (score 2) ().With new modified dual color scoring system, the chance of missing malignant lesion was 2.1%. A similar finding was observed by Shamla et al, who reported a sensitivity of 85.3%, specificity of 87.8%, PPV of 92.7%, NPV of 76.5%, and overall diagnostic accuracy of 86.2%.¹³

In current study, the mean strain ratio of 3.2 was taken as cut off to be the differentiator between benign and malignant lesions. In the 100 cases, 55 had strain ratio values less than 3.2 and 45 had more than 3.2. There was benign case of granulomatous mastitis that had shown the mean strain ratio of 7.8 and was classified under malignant category. None of the malignant lesion was classified as benign and thus it helps in avoiding unnecessary breast biopsy. The strain elastography ratio can also be used to differentiate the benign and malignant lesions. Malignant lesions show large aggregates of elastin fibres which is called elastosis. Apart from elastosis malignant cells also change the extracellular matrix and increase the cellularity thus malignant lesions are stiffer than the benign ones¹⁴.Ozel and Ozel took 3.1 as the cut off ratio which is similar to the current study¹⁰. Bojanic et al, obtained a higher strain elastography cut off of 3.8 to differentiate the benign and malignant lesions and had the sensitivity and specificity of 90.5% and 93.2 % respectively¹⁴.

Limitations

The study had certain limitations. The sample size was comparatively small to give the relevant data for the rare conditions like phyllodestumor, granulomatous mastitis and squamous cell carcinoma. The expertise of the operator plays the important role in the diagnosis.

Elastography can be the good adjuvant to conventional B mode ultrasonography to differentiate between benign from malignant lesion thus by reducing the number of biopsies. Ultrasound with cut off ratio, modified and new modified color scores increase the sensitivity, specificity, positive predictive value, negative predictive value and diagnostic accuracy against the ulrasonography alone. Elastography also plays important role in characterising the diagnosis as compared to stand alone use of USG. Further studies are required to explore the role of elastography parameters in the real time elastography and elastography guided breast biopsies to determine its role in breast malignancy screening programs.

1. Malvia S, Bagadi SA, Dubey US, Saxena S: Epidemiology of breast cancer in Indian women. Asia Pac J ClinOncol 2017;13(4):289–295.
2. Tanter M, Bercoff J, Athanasiou A, Deffieux T, Gennisson JL, Montaldo G, et al. Quantitative assessment of breast lesion viscoelasticity: initial clinical results using supersonic shear imaging. Ultrasound Med Biol 2008;34:1373-1386
3. Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic moduli of breast and prostate tissues under compression. Ultrason Imaging 1998;20:260-274.
4. MacSween RN, Burt AD. Histologic spectrum of alcoholic liver disease. Semin Liver Dis 1986;6:221-232.
5. Barr RG: Principles of elastography, in: Breast Elastography. 1st ed. Thieme Medical Publisher, 2
6. Frank DN, Robertson CE, Hamm CM, Kpadeh Z, Zhang T, Chen H, et al. Disease phenotype and genotype are associated with shifts in intestinal-associated microbiota in inflammatory bowel diseases. Inflamm Bowel Dis 2011;17:179-184
7. Menezes R, Sardessai S, Furtado R, Sardessai M: Correlation of strain elastography with conventional sonography and FNAC/Biopsy. J ClinDiagn Res 2016;10(7):TC05–10.