A pilot evaluation of magnetic resonance imaging characteristics seen with solid papillary carcinomas of the breast in 4 patients

Papillary carcinomas constitute 1–2% of breast carcinomas in women. Solid papillary carcinoma (SPC) is a rare variant of papillary carcinoma with unique pathological morphology and biological behavior [1, 2] and has recently been classified as a new category of breast papillary carcinoma by the World Health Organization (2012), differentiating it from the previous classification as a type of intraductal papillary carcinoma [1, 3‐5]. Although still under investigation, the prognosis of SPC seems to be better than that of intraductal papillary carcinoma (the most common papillary tumor) because of SPC’s unique pathological pattern [1‐3].

Pathologically, SPC is characterized by round, well-defined nodules composed of densely low-grade ductal cells separated by fibrovascular cores, leading to a morphologically solid growth pattern at low magnification [1, 2]. An underlying fibrovascular stromal network and a solid morphologic appearance are typically observed for SPC. This is in contrast to the papillary fronds covered by stratified columnar cells with uniform hyperchromatic nuclei seen in intraductal papillary carcinoma [1, 3‐5]. For Immunohistochemistry (IHC) analysis, neuroendocrine differentiation is commonly presented in SPC, but not in intraductal papillary carcinoma [1, 3].

MRI is widely used to examine breast lesions, including many varieties of papillary tumors [4‐8]. However, there is only one case report on T₁-MRI of SPC [9]. The paucity of SPC imaging research is likely due to the relatively new classification of SPC as a unique papillary tumor. Therefore, this study aims to fill the gap in the current MRI knowledge of SPC by examining this papillary carcinoma using multiple MR image modalities including T₁W and T₂W MRI, DCE-MRI, DWI and MRS.

The age of the four patients ranged from 66 years old to 79 years old. Clinical findings including patient age, location, bloody nipple discharge, and duct ectasia are listed in Table 2. For cases 1, 3, and 4, all lesions were located in the medial quadrant of the breast. For case 2, the lesion was located in the lateral quadrant of the breast. For cases 1 and 4, patients had bloody nipple discharge accompanied by duct ectasia. For cases 2 and 3, the patients showed no bloody nipple discharge and had no dilated ducts. Histography and IHC results are shown in Fig. 1 a-d. Case 2 and 4 were associated with a mucinous component on pathology. No lymph node metastasis was found in any case, and all cases showed no evidence of metastasis or recurrence at one-year follow up.

The maximum lesion sizes ranged from 0.8 cm to 3.2 cm among 4 patients. Features of the lesions extracted from conventional MRI are listed in Table 3. All lesions showed hypo-iso signal intensities on T₁ FSPGR sequence (Fig. 2a, 3a, 4a), and iso-hyper signal intensities on T₂WI and STIR sequence (Fig. 2b, 3b, 4b, 5a). On DCE-MRI, all lesions showed mass enhancement with oval (cases 1 and 2, Fig. 2c, 4c) and irregular (cases 3 and 4, Fig. 3f, 5f) shapes, and margins were circumscribed (Fig. 2c, 4c). Furthermore, internal enhancement was homogeneous in cases 1 (Fig. 2c) and heterogeneous in cases 2,3 and 4 (Fig. 4c). TIC showed rapid initial enhancement (90s) followed by plateau delayed enhancement in cases 1 and 2 (Fig. 2e), while rapid initial enhancement with washout delayed enhancement was observed in cases 3 and 4 (Figs. 3d, 5c).

All lesions on DWI (b = 1000s/mm²) were slightly hyperintense with ADC values ranging from 1.3 × 10⁻³ mm²/s to 1.9 × 10⁻³ mm²/s (Fig. 2d, 3c, 4d, 5b). Cho peak was absent at 3.2 ppm in all cases (Fig. 3e, 4e, 5d and e).

In our study, all lesions showed iso-hypo signal intensities on T₁ FSPGR, and iso-hyper signal intensities on T₂WI or STIR sequences, which included hemorrhagic or mucus components microscopically regardless of lesion size. To date, only one report by Yoshimura et al. [9] pointed out diffuse nodules surrounding the tumors in bilateral SPCs based on post-contrast T₁ MRI.

For DCE-MRI, mass enhancement was characterized with either oval or irregular shapes, and all lesion margins were circumscribed. Furthermore, internal enhancement was homogeneous or heterogeneous. In the TIC assessment, signal intensity increased rapidly in the initial contrast phases of all lesions, which is in accordance with the TIC characteristics of most breast cancers (rapid initial enhancement) [5, 6, 11]. Meanwhile, persistent or washout type of TIC appeared in delayed contrast phases of all four cases. All SPCs features observed from DCE-MRI are in line with findings in other papillary tumors [4‐7, 12, 13].

DWI is widely used to differentiate benign lesions from malignant lesions based on ADC values [4, 14, 15]. Typically, if ADC value is higher than the threshold of 1.2 × 10⁻³ mm²/s, the lesion is considered to be benign based on reports in the literature [11, 14]. However there are many exceptions because of the heterogeneity of tumor nature [15, 16]. In this study, ADC values of all SPC lesions were found in the range of 1.3–1.9 × 10⁻³ mm²/s, which is higher than the threshold. We speculate that high ADC value of SPC is due to the cystic or mucus components in the cell structures of SPC. It has been demonstrated that ADC values in mucinous cancer are higher than the threshold due to relatively free motion of water molecules in the mucin pool [16]. To further investigate whether high ADC value is a characteristic of SPC, a study including more SPC cases is underway in our institution.

The MRS results presented in this paper are not only the results of the first MRS study for SPC but also for breast papillary tumors. MRS has frequently been used to help differentiate benign breast lesions from malignant breast lesions based on the absence of signals from choline-containing compounds (Cho) [10]. Cho is a measure of increased cellular turnover and thus is generally elevated in tumors, particularly for high grade tumors. Previous studies have demonstrated that Cho signal could be low or even absent for low-grade breast tumors [10, 17, 18]. In this study, Cho peak was absent in all four cases, which suggests that the concentration of choline-containing compounds in these SPCs may be too low to be detected. The absence of Cho signal could be due to the low-grade nature of SPC in situ which grow from intervening stromal cells [18, 19]. The combination of the absence of Cho peak with a high ADC value could be a useful index to distinguish SPC in situ from other types of invasive breast carcinomas and worth further investigation with more SPC cases. Therefore, a careful study including investigating ADC values and Cho peak presentation for SPC is underway in our institution.

In our study, all patients were postmenopausal women. This finding is in accordance with previous studies which revealed that SPC primarily affects elder women [20, 21]. All lesions in our study were unilateral with no lymphadenopathy, in accordance with the majority of reports [6, 9]. Lesion size varied from 0.8 cm to 3.2 cm, which is in the range of current literature reports (1 cm to 15 cm) [6, 9, 20, 21]. However, our patients with bloody nipple discharge and accompanying dilated ducts all had lesions that were far away from the nipple. These findings differ from previous reports that SPC lesions always arise in the central (retro-areolar) area of the breast with no specific clinical features at presentation [6, 9, 22].

In conclusion, SPC MRI is characterized by heterogeneous signal intensity within the lesion, mass enhancement with circumscribed margins, either oval or irregular shapes, and the presence of a rapid enhancement in initial contrast phases. Moreover, high ADC values and the absence of Cho peak may provide valuable information for distinguishing SPC from other invasive breast carcinomas.