CD34+ Fibrozyten im Stroma benigner und maligner Tumoren der Glandula parotidea

Background: Little is known about the distribution and incidence of CD34+ fibrocytes in the stroma of parotid glands. Therefore, the present study was designed to determine the question whether the distribution of stromal CD34+ fibrocytes and α-SMA+ myofibroblasts differs between the stroma of healthy parotid glands as well as benign and malignant lesions of the parotid gland. Method: In the present study, a retrospective immunohistochemical analysis was undertaken on a study population of 122 cases in total – comprising 8 cases of normal parotid gland tissue, 68 benign and 46 malignant neoplasms of the parotid gland – from a monocentric register, dated 2000-2009, at the Institute for Pathology at the University Hospital of Marburg. Additional to a semi-quantitative analysis in accordance with a defined evaluation mode and follow-up descriptive statistics, an explorative analysis was performed, whereby the data gathered from the 3 sub-groups were evaluated globally as well as through individual group comparisons, using Fisher’s Exact Test. Results: In the stroma of normal parotid gland a loose, reticular meshwork of CD34+ fibrocytes is displayed. The stromata of healthy parotid glands do not show any α-SMA+ myofibroblasts. The stromata of pleomorphic adenomas, warthin tumours, nonmembranous basal cell adenoma and non-sclerotic oncocytoma do not exhibit any CD34+ fibrocytes. In the benign tumour group, the membranous basal cell adenomas as well as the sclerotic oncocytoma have a special position, since they have a homogenously distributed stroma network of CD34+ fibrocytes. With exception of a few warthin tumours, no additional α-SMA+ myofibroblasts could be shown in the stroma of benign parotid gland neoplasms. The consistent occurrence of CD34+ fibrocytes in the capsule of benign parotid gland tumours may act as a pathognostic characteristic. The absence of CD34+ fibrocytes in the stroma of benign parotid tumours should not be classified as criteria of malignancy in histopathological diagnostics. In almost all of the examined cases, the stroma of primary and secondary parotid carcinomas exhibited a complete, stereotypical loss of CD34+ fibrocytes, whereas the surrounding tumour-free parotid stroma revealed a standard network of resident CD34+ fibrocytes. Simultaneously, an increase of stromal α-SMA+ SPARC+ myofibroblasts could be determined for the majority of the carcinoma. This transformation of stroma cells from the CD34+ α-SMA- SPARC- to the CD34- α-SMA+ SPARC+ phenotype seems to take place in a preexisting cell population of autochthonous CD34+ fibrocytes. The loss of CD34+ fibrocytes and the concomitant appearance of α-SMA+ myofibroblasts may be applied as an additional tool to differentiate benign from malignant neoplasia in histopathological diagnostics, but should never be consulted as a single criterion of dignity. In terms of phenotyping the parotid gland stroma, it could be shown globally that there are clear differences in the distribution of CD34+ fibrocytes and α-SMA+ myofibroblasts between healthy parotid glands as well as benign and also malignant parotid tumours. Consequently, an immunophenotypic characterisation of the respective stroma is possible. Adenoid cystic carcinoma represents a diagnostic challenge, since the phenotype transformation of the stroma cells is absent. The stroma shows a preserved population of CD34+ fibrocytes, whereas α-SMA+ myofibroblasts were not detectable. This finding implies that this highly aggressive carcinoma has a different biological behaviour. Therefore and in contrast to all the other parotid carcinoma examined, a preserved CD34+ cell population in the stroma of adenoid cystic carcinoma does not exclude malignancy. Furthermore, the presence of CD34+ Fibrocytes in the stroma of adenoid cystic carcinoma may indicate malignancy. Summary: This study on 114 tumours of the parotid gland shows for the first time comparable data, in terms of distribution differences of CD34+ fibrocytes in parotid gland tumours. Stromal remodelling associated with invasive carcinomas of the parotid gland is characterised by a loss of CD34+ fibrocytes paralleled by a gain of α-SMA+ myofibroblasts and increased expression of SPARC except adenoid cystic carcinoma. The diagnostic impact of this finding is that CD34+ fibrocytes and α-SMA+ myofibroblasts may be helpful tools in distinguishing benign parotid tumours from invasive parotid cancer. The lack of CD34+ fibrocytes in the stroma of parotid gland tumours should never be used as a single criterion of dignity. SPARC expression appears to be closely related to malignancy in parotid tumours. In conclusion, an additional prospective, multicentre study is needed to confirm the results of this explorative, retrospective study and to validate the diagnostic potential of CD34+ Fibrocytes.