The prognosis of the current PC patient might unusually appear better than that reported by Silver and Tavassoli
, since it is likely that the final diagnosis was made at a relatively earlier stage of the tumors; even though follow-up period was not too long in this limited case report. Actually, our case has shown neither local recurrence nor metastases for 10 months after the operation. In contrast, Nguyen et al. recently have described that, although not all of these PCs behave badly, the presence of spindle cell sarcomatoid (metaplastic) components or the tumor size larger than 5 cm significantly leads to confer a significantly poor prognosis even in the early stage for PC of the breast
. In addition, as shown in this PC case, no immunohistochemical expression of E-cadherin and β-catenin, involved in cell adhesion and cell-cell interaction, very likely results in severe vessel permeation, advanced clinical stage, and worse outcome
[11, 12]. Hence, it would be critical for very early treatment against PC to establish an accurate preoperative diagnosis by fine needle aspiration cytology, the clinical utility of which in diagnosing breast tumors has been already generalized. The cytological characteristic features of PC of the breast seem to reflect the histopathological ones, exhibiting numerous individual bizarre and giant malignant cells with pleomorphic irregular nuclei, coarse chromatin, conspicuous nucleoli, very high mitotic rate, and relatively abundant cytoplasm in the background of possible inflammation and/or necrosis
[6–8]. Indeed, when extensive necrosis is present without evidence of viable PC cells on cytology specimens, cytopathologists should exclude out the possibility of benign breast tumors, e.g., infarcted fibroadenoma
. On the other hand, we for the first time indicate that cohesive, three-dimensional and/or sheet-like clusters of highly malignant cells coexist on PC cytology. In fact, the cytological findings of this relatively new and extremely rare entity have never been well described or reviewed more recently. The cytodiagnostics might not need to come out with such a distinctive one, but should confirm malignancy in any case of PC of the breast. Despite that, a confident and accurate diagnosis of PC might be possible only on cytology specimens, owing to its cytomorphologically peculiar characteristics, adequate samplings, and/or accumulated experience. Furthermore, few previous papers proposed that, when the cytologic features of invasive ductal carcinoma with bizarre pleomorphic malignant giant cells were difficult to make an accurate and conclusive diagnosis, immunostaining for cytokeratins and EMA on cell blocks of aspirates would be very useful for the diagnosis of mammary PC
[6, 8]. Nevertheless, in cases with evidence of single bizarre giant cells and clusters of them, as shown here, cytopathologists should raise high possibility of PC as one of differential diagnoses, at the very least. Future studies are further needed.
It is possible that the present case report might be pathologically remarkable for three reasons at least: first, cystic cavity formation was uniquely recognized within the tissue of PC. Amongst the ‘true’ cases reported or retrospectively considered as PC of the breast in the English literatures
[1, 3, 6–8, 10, 14], several PC tumors have displayed nodular or mass lesions with cystic change or cystic cavity formation
[6, 8, 14], similar to the present case. We can hypothesize that the cystic cavity of mammary carcinoma would be formed via multiple processes: carcinoma cells initially develop in ductal wall and locally (i.e., intraductally) grow up in volume, and next block the ductal lumen possibly via their stromal involvements; and finally, the cystic cavity might gradually get larger with increased inner pressure and/or possible necrotic change. However, this hypothesis seems to be too speculative and not to be based on the histopathological features of our case. Nevertheless, since it remains to be elucidated whether PCs of the breast are prone to having cyst formation, it would be intriguing to study this topic.
Second, we have found the presence of ‘emperipolesis’ wherein surrounding inflammatory cells are engulfed by tumor giant cells of PC, very similar to pleomorphic giant cell carcinomas of the lung
. According to the recent, relatively large study of PCs of the breast by Nguyen et al., many cases of them showed neoplastic ‘cannibalism’ of their tumor cells, but not of inflammatory cells. However, future thorough studies also will be further required to determine whether these features are genuinely characteristic in mammary PCs after histologically examining a larger number of PC cases.
Third, immunohistochemical analyses of not merely epithelial markers, including CK5/6 and 34βE12, but CD10 and vimentin were positively expressed in the tumor nests. Additionally, ER/PgR/HER2 were all triple-negative, possibly corresponding to basal-like type breast cancer
. Although there have been no large, detailed immunohistochemical studies of PC of the breast, to date, our data imply that those carcinoma cells might have potential squamous, myoepithelial, and/or basal-like phenotypes, and epithelial-mesenchymal transition (EMT), as well, supported by some published papers
[14, 16]. We might provide the possible evidence that PCs arise from ductal epithelial-myoepithelial cells, as a result of neoplastic transformation of outer supporting myoepithelial cells, as well as inner ductal epithelial cells
[17, 18]. However, since other myoepithelial markers examined, such as α-SMA, S-100 protein, p63, and calponin, were completely negative, this implication would be highly speculative and unlikely. First of all, pathological differential diagnosis of this mammary PC case could be metaplastic carcinoma, and indeed, metaplastic carcinoma of the breast include squamous cell carcinomas, spindle cell carcinomas, or carcinomas with mesenchymal differentiation
. In this context, not only by definition but from the above immunohistochemistry, it is very likely that PCs with metaplastic spindle and/or squamous differentiation are metaplastic carcinomas per se. Some confusion still exists in the newest WHO classification
, and thus it could be very challenging that we pathologists strictly make a final diagnosis as PC of the breast.
It is possible that, based on these characteristic clinical and/or cytopathological features, as described above, PC of the breast might be a special new entity of the breast cancer, but not a rare variant of invasive carcinoma of no special type. In this context, the present case report could interest the scientific community, taken together with new cytological findings and specific recommendations for cytodiagnostics.