BACKGROUNDDue to the low incidence of salivary duct carcinoma (SDC), only limited data in regard to the biologic behavior of this tumor and its immunohistochemical characteristics are available. The authors analyzed the clinical, molecular, and genetic profile of SDC and identified prognostic factors.METHODSThe follow‐up of 50 patients with SDC was obtained and paraffin‐embedded tumor samples were examined immunohistochemically. In all samples, the expression of Ki‐67, HER‐2, and the oncoproteins p16 and p53 was examined immunohistochemically, followed by a mutation analysis of p16 and p53. The survival rate was calculated by the Kaplan–Meier method and prognostic variables were analyzed with the log‐rank test.RESULTSSDC predominantly effected male patients (66%) in their 7th decade of life. SDC mainly occurred in the parotid gland (78%; submandibular gland, 12%; minor salivary glands, 10%). Approximately two‐thirds of the patients (33 of 50) presented with a T3/T4 tumor. In 28 patients (56%), cervical lymph node metastasis was present at the time of diagnosis. Local disease recurrence was observed in 48% of patients an average of 17.4 months after initial treatment. Distant disease metastasis developed in 48% of patients an average of 29 months after initial treatment. The average overall survival period was 56.2 months. In the current study, 20.6% of the probes with positive HER‐2/neu expression were (+++) positive. p53 was expressed in 83.9% of the tumor samples. In 11.8% of the tumor samples, there was a lack of p16 expression.CONCLUSIONSMutations of the p53 gene were more frequent in tumor samples with (++) and (+++) immunoreactivity and mainly affected exons 7 and 8. A mutation of the p16 gene was only found in 1 tumor sample. Expression of HER‐2/neu and p53 was statistically linked (P < 0.05) to early local disease recurrence, distant disease metastasis, and survival rates. Cancer 2005. © 2005 American Cancer Society.

No abstract

No abstract

The diagnosis of salivary gland myoepithelioma, an entity with heterogeneous cytomorphology and inconsistent immunophenotype, rests on conventional histology. However, the clinical course cannot be predicted reliably from cytomorphological and immunophenotypic analysis. The present study determined the immunophenotype of a representative series of 12 myoepitheliomas and 21 malignant myoepitheliomas. Among the seven markers tested, antibodies against cytokeratins 5/6, S-100 protein, and vimentin produced the most consistent reactivity profile. Comparative genomic hybridization (CGH) profiles of 12 myoepitheliomas showed chromosomal losses in three of 12 cases. In myoepithelial carcinomas, however, ten of 19 tissues investigated by CGH lacked detectable cytogenetic aberrations. In five cases, aberrations involved chromosome 8, in line with observations in salivary gland carcinomas of other differentiation. One case that was represented in three separately localized manifestations of the disease proved informative as to the relevance of gross aberration for tumour development, as these tumours differed in their CGH profiles. Staining for cytokeratins 5/6 is a useful addition to the established immunohistological marker panel in the work-up of myoepitheliomas, because of its reliable expression in most cases and because it may underline the epithelial nature of the lesion. CGH proved to be of limited value as a diagnostic adjunct; the presence of numerous gross cytogenetic aberrations should raise the suspicion of malignancy. The low frequency of aberrations detectable by CGH in overtly malignant myoepithelial neoplasms suggests that gross cytogenetic alterations were acquired in the course of tumour progression and points to the relevance of genetic changes not resolved by CGH.