The term
lymphoepithelioma was introduced in 1921 to refer to an undifferentiated carcinoma of the nasopharynx with a dense lymphocytic component [
33]. It is characterized by nests of undifferentiated epithelial cells infiltrated by a prominent benign reactive lymphocytic infiltrate. Epstein-Barr virus is frequently present in the malignant epithelial cells of this carcinoma [
17]. Such association was first reported by Harold Zur Hausen in 1970 [
19]. LELC is a subtype of poorly differentiated squamous cell carcinoma, identical to the LE described in the nasopharynx but originating in other anatomic sites. Most of the reported cases of primary LELC of the salivary gland occurred in Asians and Eskimos whereas they are uncommon in Americans and Europeans [
26]. To the best of our knowledge, 14 cases have been previously described in western people [
1,
4,
25‐
32]. All lesions affected parotid gland. The median age of the patients was 54,6 years (range: 32–74), with a female predominance (78.5%). Six patients (43%) showed lymph node metastasis (not all data available). Only two patient (14%) received radiation therapy alone, the others being treated by total parotidectomy with or without postoperative radiation therapy. All the patients were alive at the last follow-up (not all data available).
In situ hybridization for EBV was positive only in 5 patients (36%, not all data available). Clinico-pathological and therapy data of all these 14 cases are listed in Table
1.
Table 1
Clinicopathological features of patients with LELC of the parotid gland described in the Western literature
| F | 36 | SP | At the last follow-up, no signs of relapse | n.a. | Cervical lymph node metastasis |
| F | 55 | SP | At the last follow-up, no signs of relapse | n.a. | n.a. |
| F | 32 | RP | At the last follow-up, no signs of relapse | n.a. | n.a. |
| M | 51 | RT | At 11-year follow-up, no signs of relapse | n.a. | Cervical lymph node metastasis |
| F | 42 | SP + RT | At 12-month follow-up, no signs of relapse | n.a. | n.a. |
| F | 68 | SP + RT | At 24-month follow-up, no signs of relapse | negative | Metastasis to intraparotid lymph node |
| F | 57 | RP + RT | At 24-month follow-up, no signs of relapse | negative | Lymph node metastasis |
| M | 64 | SP | At the last follow-up, no signs of relapse | positive | Regional lymph node metastasis |
| F | 45 | SP + RT | At 15-month follow-up, no signs of relapse | positive | n.a. |
| F | 54 | RP + RT | At 24-month follow-up, no signs of relapse | positive | n.a. |
| F | 72 | SP + RT | At 36-month follow-up, no signs of relapse | positive | Periparotid lymph node metastasis |
| F | 74 | RT | n.a. | positive | n.a. |
| M | 47 | RP + RT | At 7-month follow-up, no signs of relapse | negative | Neural metastasis |
| F | 67 | SP + RT | At 12-month follow-up, no signs of relapse | n.a. | n.a. |
The origin and pathogenesis of LELC is still unknown [
1]. It is thought to arise in either of two settings: malignant transformation of a myoepithelial island or malignant transformation of glandular and ductal inclusions in intraparotid lymph node. The association with EBV has suggested a possible role for the virus in the etiology but the relationship is controversial, considering that some LELC at specific anatomical sites have never been proven to be associated with EBV [
17]. There is no satisfactory explanation for EBV being commonly present in certain anatomic sites, but not in others. In the past, it was believed that only foregut-derived organs (salivary gland, stomach, thymus, and lung) were susceptible to EBV-associated carcinogenesis, perhaps because these organs are in close proximity to sites of natural viral replication [
19]. The portal of entry of the virus is thought to be the oropharyngeal mucosa, which also serves as the site of production of the virus, which is periodically shed in saliva [
27]. The infection of lymphocytes permits systemic dissemination, and following primary infection, EBV lies latent in a few lymphocytes for the duration of life. The establishment of a persistent infection protects the virus against the immune response. Thus, EBV can immortalize human keratinocytes promoting neoplastic transformation. The oncogenic role of EBV is elicited by its products. Among these, LMP-1 has a driving role. Expression of LMP-1 prevents apoptosis, induces abnormal cell proliferation, deregulates epithelial growths and inhibits differentiation, with the epithelial cells showing the features of transformed cells [
34]. However, the expression of this protein was observed only in few EBV-linked salivary LELC from Asian patients and our case is the first one in which LMP-1 positivity was detected in a Western patient. These findings may suggest that different strains of EBV could be involved in the pathogenesis of salivary LELC in different geographic areas. These strands may express LMP-1 with different genetic sequence and thus different immunohistochemical positivity [
35]. A recent study has reported the immunoreactivity of LELC neoplastic cells for fascin as in our case [
4]. Fascin is a 55 Kda globular protein that is expressed in almost all cellular types. It acts organizing actin-based structures (i.e. filopodia, microspikes and lamellipodial ribs), and cytoplasmic microfilament bundles [
36,
37]. In addition fascin packs F-actin into parallel bundles. The association of fascin with F-actin is strongly regulated by the extracellular matrix (ECM) environment of cells and is required for cell migration [
38,
39]. It is possible that latent EBV infection of epithelial salivary gland cells can upregulate transcriptional activity of the fascin-1 gene with increased synthesis of the protein in cytoplasm. This phenomenon could play an important role in the progression to an invasive phenotype of transformed cells, with the loss of cell to cell adhesion and loss of junctional communications. Furthermore, fascin-positive stromal dentritic cells might be modified by functional changes via the EBV induced release of soluble CD83 with inhibition of stimulation of T-cell proliferation [
39]. However, the absence of EBV genome in most LELC cases [
19] implies that EBV is not a necessary factor in the aetiology or pathogenesis of LELC, and that genetic, environmental or geographic factors may be involved [
17].