Breast Involvement by Hematologic Malignancies: Ultrasound and Elastography Findings with Clinical Outcomes

INTRODUCTION

Hematological malignancies very rarely involve the breast.[1] Lymphoma and leukemia are the most common hematological malignancies affecting the breast.[2] The published data on clinical and radiological features of breast leukemia and lymphoma depends predominantly on case reports or case series. Mammography, magnetic resonance imaging (MRI) and B-mode ultrasound (US) findings of breast leukemia and lymphoma have already been reported.[1345678] However, elastography (ES) findings of breast involvement by hematological diseases have rarely been reported.[91011] Gkali et al., reported a case of primary non-Hodgkin's lymphoma of the breast that was soft, and Barr et al., reported a case of breast lymphoma with a soft lesion on ES.[910] Then, Barr et al., also reported four cases of primary breast lymphoma that presented with soft lesions.[11]

ES is a recently developed US technique enabling the evaluation of tissue stiffness. The American College of Radiology includes elasticity assessment in the fifth edition of Ultrasonographic Breast Imaging Reporting and Data System (BI-RADS US).[12] Recently, there is an increasing number of studies assessing the role of ES in detecting breast malignancies, especially primary breast carcinoma.[13141516] In addition, ES has been shown to provide additional information in differentiating between benign and malignant breast lesions, and determining biopsy. Strain and Shear Wave ES are two different ES techniques. It has been reported that Shear Wave ES is a more objective method than strain ES. Shear wave ES enables quantitative measurement of tissue stiffness.[17] In this study, we aimed to analyze B-mode US and ES findings of patients with clinical outcomes.

MATERIALS AND METHODS

Our Institutional Review Board approved this retrospective study and waived the informed consent. We searched all core-needle biopsy results performed at our breast center between the dates of January 2013 and September 2016. Our search revealed 9 patients having breast involvement due to hematological malignancies. All the diagnoses were confirmed using US-guided core-needle biopsy. One patient had strain ES and 8 patients had Shear Wave ES images. B-mode US imaging findings and ES findings were analyzed. In addition, clinical outcomes of the patients were noted.

Real-time whole B-mode breast US examinations and ES were performed on all patients using ES or Shear-Wave ES. We used two US systems (Acuson S 2000, Siemens, Erlangen, Germany) one of which only enables strain ES and the second enables Shear-Wave ES. A linear transducer that enables scanning with a frequency range of 4 − 9MHz was used. Two radiologists reviewed the cases at the same time on consensus and conferred on every case. There was a case report form, and the radiologists were requested to decide the pattern of the infiltration (diffuse, solitary or multiple mass), determine the BI-RADS categorization of the mass, and describe both US and ES findings together. The elasticity assessments were done by general impression of color-coded maps, quantitative values (including strain ratio), and Shear-Wave Velocity (SWV). Virtual touch imaging (VTI) maps were also used when available.

Statistical analysis was performed using the statistical package SPSS software (version 17.0, SPSS Inc., IL, USA). If continuous variables were normal, they were described as mean ± standard deviation (P > 0.05 in Kolmogorov–Smirnov test or Shapira–Wilk [n < 30]), and if they were not normal, they were described as the median.

RESULTS

Patient characteristics, B-mode US, and ES findings and clinical outcomes are summarized in Table 1. The mean age of the study population was 46.6 years (range, 20–83 years). All patients were female. One patient had primary breast lymphoma whereas eight of the patients had secondary involvement of breast with leukemia or lymphoma. B-mode US images were available for all patients. One patient was examined using strain ES and 8 patients were examined using Shear-Wave ES. Three patients had bilateral, and 6 patients had unilateral involvement. None of the lesions had calcifications. Two patients had diffuse infiltration and 7 patients had mass lesions.

Table 1 Patient characteristics, B-mode ultrasound and elastography findings and clinical outcomes

Diffuse infiltration (n = 2)

Two patients that had diffused parenchymal infiltration had similar B-mode US findings including diffuse hypoechoic finger-like infiltration of the normal breast parenchyma [Figure 1]. One patient had bilateral diffuse and one had left-sided diffuse parenchymal infiltration. One of these patients had accompanying skin edema that can be detected through US. These infiltrations had no posterior acoustic features. The elasticity assessments of these lesions were soft and intermediate. The patients with diffuse infiltration of the breast had a poor prognosis. One of them had already passed away, and the other had a progressive disease that was being treated.

Close

Figure 1

57 year's old female with a diagnosis of peripheral T-cell lymphoma not otherwise specified (a) B-mode Ultrasound showed a well-circumscribed round shaped hyperechoic lesion including centrally hypoechoic areas within the mass. (b) Shear Wave elastography demonstrates Shear Wave velocities from 3.52 to 5.23 m/s within the mass. (c) Virtual Touch Imaging method demonstrates diffuse dark lesion with same size as B-mode ultrasound.

Export to PPT

Mass lesions (n = 7)

The size of the tumors varied from 0.8 to 7.0 cm, which was the largest diameter (median, 2.5 cm.). Two patients had a unilateral solitary mass. Five patients had more than one lesion, either unilateral or bilateral. Three of these seven masses had circumscribed, and the remaining 4 had uncircumscribed margins [Figure 2]. The BI-RADs scores of the patients were as follows: BIRADS 3 (n = 1), BIRADS 4A (n = 2), BIRADS 4B (n = 2), and BIRADS 4C (n = 2). The elasticity assessments of the patients were as follows: soft (n = 1), intermediate (n = 4), high (n = 2). SWVs are summarized in Table 1. Five of these patients were examined by grayscale VTI maps. VTI maps showed that, in 3 of these lesions, the lesion was diffuse dark and larger than B-mode. Using VTI, two patients were found to have diffuse dark lesions, that were on the same size compared with B-mode US.

Close

Figure 2

33 year's old female with a diagnosis of non-Hodgkin lymphoma. (a) B-mode ultrasound shows diffuse hypoechoic infiltration of the normal breast parenchyma. (b) Shear Wave elastography demonstrates Shear Wave velocities from 1.92 to 2.43 m/s. (c) Virtual touch imaging method demonstrates diffuse dark lesion.

Export to PPT

DISCUSSION

Breast manifestations of hematological malignancies are rare.[3] Radiologic features of breast leukemia and lymphoma have been reported only sporadically through case reports and have been nonspecific.[12] Mammography reveals masses and architectural distortions. US shows mainly hypoechoic, microlobulated, or indistinct masses. MRI showed nonspecific imaging findings, including single or multiple enhancing masses or nonmass like enhancing lesions, and restricted diffusion.[4578]

A prior history of hematological malignancies may be helpful in diagnosing secondary breast lesions; however, the breast manifestation may be the initial diagnosis of systemic diseases.[18] Another reason to differentiate primary breast cancer from hematological involvement among the patients with hematological malignancies is because it affects both treatment and clinical outcome of the disease.

The purpose of this article is to describe the US and ES findings of hematological malignancies of the breast and assess the correlations between imaging findings and clinical outcomes.

In our study population, both lymphoma and leukemia demonstrated diffuse parenchymal infiltration. On the other hand, mass lesions were more common than diffuse parenchymal involvement among our study population. Mass lesions showed both circumscribed and uncircumscribed margins. The hypoechoic echo pattern was the most common echo pattern. Mass lesions manifested as either a solitary mass or multiple masses. In general, higher strain ratios and SWVs suggested malignant lesions in the breast. In our study, mass lesions showed elasticity assessments from soft to high. High elasticity scores may alert us to further investigate the lesions. In the literature, different cutoff SWVs-ranging from 4.1 to 5.2 m/s-were reported to differentiate malignant lesions from benign ones.[10192021] However, the potential risk is the lesions with low elasticity assessments. Previously reported studies showed the potential role of quantitative and qualitative ES in affecting the BI-RADS score.[2223] In our study, only one patient was categorized as having BI-RADS 3; however, the majority of the patients were considered as having BI-RADS 4 lesions.

Although the follow-up period of the patients was short, the prognosis seemed to be poor with breast involvement by hematological malignancies. Both soft and hard lesions showed poor prognosis, so the prognosis appeared to be poor regardless of elasticity scores. ES findings may not be a predictive value in assessing the treatment response or survival rate. Previously, regarding breast involvement with lymphoma, it was reported that the prognosis appeared to be mainly related to age, stage of the disease, and the histological type of the disease.[24]

The limitations of the current study were the relatively limited number of patients, the nonblinded image evaluation, and the retrospective nature of the study. Short-term follow-up is another limitation.

CONCLUSIONS

We reported 9 patients with breast involvement due to lymphoma and leukemia. Both B-mode US and ES showed various imaging findings. It is important to consider that both lymphoma and leukemia may show soft or intermediate elasticity assessment. In this regard, the patients’ history and clinical background cause us to consider breast involvement in leukemia and lymphoma patients. It appears that there was no correlation between the elasticity assessment and clinical outcomes.