Cellular smears with bimodal benign pattern, numerous single bipolar stromal cell nuclei and presence of fibromyxoid stroma are the essential diagnostic features on aspiration smears from fibroadenoma of breast.[1
] The acidophilic mucopolysaccharide stromal fragments is expected to stain positive with AB. However, such staining stroma was not demonstrable in our FNA smears of fibroadenoma. The single case where myxoid stroma was present, also did not stain with AB but stained positive with PAS-D. It is suggested that this fibroadenoma with myxoid change has neutral mucopolysaccharide instead of acid mucopolysaccharide.
Certain cases of fibroadenomas and fibrocystic disease with epithelial hyperplasia may yield cellular smears with suspicious discohesion and atypia of cells suggesting low grade duct carcinoma. Intracytoplasmic PAS-D positivity though seen, is rare in fibroadenomas (only 1 out of 29 cases in our study). They were also not strickingly globular material as seen in carcinomas, but just fulfilling the criteria proposed by Johnson and Wadhera.[2
] The maximum grade that could be assigned to this weak positivity was + (the lowest). Strict criteria for recognising PAS-D-positive cytoplasmic globules need to be adhered to.
We found PAS-D-positive intracytoplasmic globules in 14/19 (74%) of our duct carcinomas. We have deliberately included more lower grade carcinomas in our study because equivocal smears with high cellularity variable but subtle discohesion and mild atypia are more likely to be confused with these in cytological smears from fibroadenomas and other benign proliferations. However, the few cases of grade II tumors included in this study showed slightly higher percentage of positivity (4/5 or 80%) as compared to 10/14 (70%) of grade I tumors. This could be the reason for slightly lower positivity (74%) in our study as compared to 90% positivity reported by Nijhawan et al
The single mucinous carcinoma encountered by us showed intracytoplasmic PAS-D-positive globules in addition to PAS-D-positive lakes of mucinous substance in the background. The quality of staining was different from that seen in the case of fibroadenoma with mucinous change, in that it was more homogenous and appeared fluid-like. Fibroadenoma also do not show intracytoplasmic globules. These differences may aid in the distinction of these two entities when fibroadenoma with mucinous change and few atypical cells are causing worry.
The PAS-D stained intracytoplasmic globules in lobular carcinoma is reported to have a distinct morphology having a dark staining central core and lighter periphery.[6
] However, we could not contribute to this observation as we had no cases of this type of carcinoma included in this study.
In pleomorphic adenoma of salivary gland, two types of mucins are demonstrable on histological sections.[7
] Our findings of PAS-D-positive material and AB-positive material in FNA smears is consistent with this observation. Luna and Pilch[8
] have also reported presence of eosinophillic colloid-like PAS-D positive material in duct lumina. Since myoepitheliomas do not yield stainable material in smears, these stains may help to identify them. Mucoepidermoid carcinoma and adenoid cystic carcinoma also needs to be distinguished from pleomorphic adenoma. Kawahara et al
] have found that the extracellular material in adenoid cystic carcinoma stains positive with AB. Hence, this stain is unlikely to be of any help in differential diagnosis. However, use of PAS-D may be useful as we have found that stromal fragments seen in some of our cases of pleomorphic adenoma stain positive with it. This feature has not been noticed in adenoid cystic carcinoma. In addition, adenoid cystic carcinoma do not show intracytoplasmic PAS-D-positive globules.
PAS-D-positive globules seen in mucoepidermoid carcinoma are of help in distinguishing it from deposits of squamous cell carcinoma. Luna et al
] emphasized this point by stating that mucin production is invariably present in mucoepidermoid carcinoma and in the absence of this, the tumor should be labelled as epidermoid carcinoma.Aspirates from Warthin tumor may yield mucin-producing cells and metaplastic squamous cells to suggest mucoepidermoid carcinoma. In this situation, mucin detection may not be of help in distinguishing these as observed by Goonewardene and Nasuti.[10
The presence of PAS-D-positive globules in FNA smears of acinic cell carcinoma have been reported.[11
] Our finding of similar globules in the single case of acinic cell carcinoma included in this study confirms this. On the smears, intracytoplasmic PAS-D positivity may be restricted to only a few cells due to the fragility and dispersal of the cytoplasm. Granular cytoplasmic PAS-D positivity is exclusive to acinic cell carcinoma. This feature can be useful in distinguishing it from clear cell tumor and oncocytic tumors.
FNA smears from basal cell adenoma showed numerous basaloid cells in clusters with some metachromatic stromal fragments in the background and hence, could be confused with pleomorphic adenoma. The stromal elements in basal cell adenoma are collagenous in origin and has no mucin.[12
] Hence, it is not expected to give positive reaction with PAS-D or AB. This is confirmed by us on sections. The PAS-D or AB-positive stromal fragments consistently present in pleomorphic adenoma help their identification when confused with monomorphic adenoma. In this scenario, the AB positivity of stroma and cytoplasmic PAS-D positive globules seen in adenoid cystic carcinoma can be reassuring in excluding basal cell adenoma.
The smears of salivary duct carcinoma had striking resemblance to those of breast ductal carcinoma. However, the cells failed to stain with either PAS-D or AB in smears or sections in this study. This is in contrast to the observations of Genester et al
] who reported presence of apocrine type PAS-D-positive cytoplasmic vacuoles in salivary duct carcinoma.This suggests the possibility of existence of mucinous variant of salivary duct carcinoma as has been reported by Simson et al
] The main differential diagnosis here is with high grade mucoepidermoid carcinoma. Since, mucin secretion is the rule in mucoepidermoid carcinoma and reports of mucin production in salivary duct carcinoma are variable, this field needs to be reported further before drawing any conclusions regarding the utility of these stains.