Phyllodes tumor is named after a leaf-like configuration of an overgrown mesenchymal structure, and is a rare disease that occurs in less than 1% of all breast tumors (
1). The phyllodes tumor is usually benign, but it may be borderline or malignant, and the percentage of malignant tumors reported in the literature varies from 8% to 45% (
2,
3). Most of the malignant phyllodes tumors show fibrosarcomatous differentiation, and other sarcomatous elements are rarely reported. Other possible sarcomatous differentiations include angiosarcoma, chondrosarcoma, and all other types of sarcomatous differentiations. Among them, liposarcomatous differentiation has been very rarely reported (
4).
On mammography, most of the phyllodes tumors were found as a round shaped, circumscribed marginated, and high density mass, with large and fast-growing tendency. A lucent halo caused by a fat component can localize the margin of the mass. Calcifications are rare because of their rapid growth (
9). In our case, the mass showed typical features on mammography five years ago, with an oval circumscribed equal density mass with a radiolucent halo. The re-emerged mass appeared similar at first but evolved as an oval indistinct high-density mass, with heterogeneous internal fat components. The differences in these mammographic findings can be interpreted as the difference between malignant potentials.
On USG, most of the phyllodes tumors exhibited a lobulated contour, smooth margins, mild hypoechoic internal echo, and heterogeneous internal echo pattern (
5). According to Liberman et al. most of phyllodes tumors exhibited hypoechoic internal echotexture and posterior acoustic enhancement and the presence of cystic spaces on USG was known to be more common in malignant cases (
6). According to Kalambo et al. an irregular shape, non-circumscribed margins, and/or mass size > 7 cm may indicate more borderline and malignant subtypes of phyllodes tumors (
7). In our case, the mass was observed as a mainly hyperechoic mass, with oval shape, indistinct margin, and cystic portions. It is noteworthy that the mass was hyper-echogenic unlike usual cases, and the mass was with cystic portions, increasing the probability of malignancy.
Most phyllodes tumors are surgically treated and full mastectomy is recommended. If the mastectomy is not performed completely or if the resection margin is positive, recurrence may occur (
10). The recurrence occurred in about 21% of all cases, most of which occurred within 2 years after surgery (
1). In our case, recurrence may be suspected, because of the positive surgical margins seen in the previous surgery and the reappearance of the tumor near the previous operation site.
According to Chao et al. of the 110 cases, almost all phyllodes tumors appeared as hypoechoic, except for three hyperechoic lesions and one isoechoic lesion. Additionally, all four non-hypoechoic tumors were benign tumors (
5). On the other hand, referring to case reports of malignant phyllodes tumors with heterologous sarcomatous differentiation, it was found that these tumors showed various echogenicity on USG, ranged from heterogeneous echogenicity to predominantly cystic mass (
11,
12). In studying these case reports, we thought that the unique USG findings found in our case report were also related to heterogeneous sarcomatous differentiation. However, there were not many case reports showing hyperechoic appearance, and only one case was found. That case presented a patient with liposarcomatous differentiation showing hyperechoic on USG, same as our case (
8). Considering these two cases together, although there is no published paper regarding the relationship between liposarcomatous differentiation and hyperechoic appearance, it can be interpreted in relation to the fact that fat is originally hyperechoic on USG.
Our case report included serial mammographies, USG images including Doppler USG, and microscope slides with pathologic explanations, compared to the previous case report, which only presented one USG image. However, since there were only two related case reports, there is a limit to the explanation of the association between liposarcomatous differentiation and hyperechoic appearance.
In conclusion, we can identify several potential rare features of various and unpredictable phyllodes tumors through our case report. First, malignant phyllodes tumors may appear as hyperechoic on USG, and second, they may show liposarcomatous differentiations, with these two factors possibly being related. In addition, phyllodes tumors can recur if they are not totally resected and they can show different malignant potentials from the previous tumor.