Introduction

Parotid gland carcinomas constitute a very heterogeneous group of cancers. Their diagnosis is made based on their morphological, cytological and biological (clinical) features [1].

Histological classification of salivary gland tumours by WHO undergoes constant changes. During the first edition dated 1972, 7 types of carcinomas were distinguished, in 1991 there were 18 of them and the currently binding edition from year 2005 lists as many as 24 types of carcinomas [2–4]. In future editions, the number of types of salivary gland carcinomas will probably increase. Introduction of new types of cancers is associated both with difficulties in qualifying them to previous categories based on their morphological features (histological) and the immunohistochemical phenotype, in specifying the criteria of diagnosis and in introducing new techniques (for example, genetic tests for confirmation of a given mutation) [5]. According to the literature, after a reassessment of histological specimens of salivary gland carcinomas, primary diagnosis may be changed in up to 1/3 of patients [6, 7]. Reclassification can involve diagnosis change not only from one type of cancer into another of the same or different grade but also from a primary to a secondary (metastasis) tumour or from a malignant to benign lesion [8]. The last two situations are especially associated with clinical implications due to a risk of overdiagnosis and overtreatment or underdiagnosis and an inadequate treatment. For this reason, the authors of this article have evaluated the relation between reclassification/verification of histological diagnosis and the clinical course of parotid gland carcinomas in a large case series.

Materials and methods

Histological and immunohistochemical reassessment of 111 specimens of parotid gland carcinomas from resections performed at the Department of Otolaryngology of Medical University of Gdańsk during the years 1992–2010 was conducted in the Department of Pathomorphology of the Medical University of Gdańsk, Poland, to verify the diagnosis and potentially reclassify the tumours. Concurrently, the same specimens were independently examined and analysed cytogenetically at the Department of Pathology of the Charles University in Prague, Faculty of Medicine in Pilsen, Czech Republic. In case of differences in the assessment of specimens between the two centres, the examination was repeated and a consensus diagnosis was established. Reclassification was carried out according to histological classification by WHO from year 2005 and with addition of some new histological types of salivary gland neoplasms described since then [4, 5, 9]. Diagnoses were changed retrospectively based on microscopic appearance as interpreted by a pathologist experienced in salivary gland neoplasms, and taking into consideration clinical/follow-up data, additional immunohistochemical (IHC) and cytogenetic tests such as fluorescent in situ hybridization (FISH). Specifically, FISH was used to confirm the presence of ETV6-NTRK3 translocations in cases of mammary analogue secretory carcinoma (MASC) and CTRC1-MAML2 and CTRC3-MAML2 translocations in mucoepidermoid carcinoma (MEC) using previously described methodology [9, 10]. The prognostic factors taken into account in evaluation of the clinical course included the stage, grade and margins status. Clinical stage was based on TNM of 2009 [11].

Results

Histological analysis

The primary diagnosis was changed in 28 of the 111 patients (25.2 %). In 16 of those patients, the change involved reclassification of the salivary carcinoma from one type to another, and specifically, in 6 cases the change was to a new type of salivary gland carcinoma not recognized in the 2005 WHO classification—mammary analogue secretory carcinoma (MASC)—based on the presence of ETV6-NTRK3 translocation. In another four cases, diagnosis was changed to salivary duct carcinoma (SDC), supported by positive expression of HER-2 protein. Two additional patients originally diagnosed with carcinoma ex pleomorphic adenoma (CxPA) on review were found to have in situ carcinoma arising in pleomorphic adenoma (CxPA in situ). In six cases, primary cancer of salivary gland was reclassified as a secondary tumour (metastases from the kidney, breast or skin), while in four other cases the diagnosis of carcinoma was changed to a benign neoplasm (adenoma) and one case to a non-neoplastic lesion (necrotizing sialometaplasia). The 28 diagnostic changes are summarized in Table 1.

Table 1 Methods and basis for reclassification of 28 parotid gland carcinomas
Full size table

Clinical analysis

Table 2 presents clinical and histopathological data of patients, whose diagnoses were revised. A potentially atypical clinical course was observed in 4 out of 28 patients, whose diagnoses were changed. In 2 of those 4 patients (no. 25 and 26), the course of the neoplastic disease was more aggressive (generalized neoplastic disease and death) than would be predicted based on original histological diagnosis, low grade, stage, and completeness of surgical excision. In both cases, the parotid gland tumour turned out to be a metastatic lesion on review. In 2 other patients (no. 17 and 21), the situation was opposite—unexpected asymptomatic course lasting for many years after tumour resection with uncertain margin. In the first case, the lesion was rediagnosed as “in situ” carcinoma ex pleomorphic adenoma and in the second case, as a non-malignant neoplasm (a rare variant of myoepithelioma).

Table 2 Clinical course and follow-up of patients with revised diagnosis
Full size table

Discussion

For many clinicians, a change of histological diagnosis represents a certain taboo. In our series, the primary histopathological diagnosis was changed in about a quarter of patients. We were able to find only a few other reports discussing this issue. Van der Wal et al. reassessed specimens of tumours of small salivary glands and of the parotid gland, which resulted in change of diagnosis in 29 and 11.7 % of patients, respectively. In that series, histological verification and reclassification was based exclusively on a repeat microscopic examination of the specimens. It is worth to point out that after histological revision, the diagnosis of 7 adenomas was changed to carcinoma (total in intraoral and parotid location): (2 polymorphous low-grade adenocarcinoma, 2 MEC, adenoid cystic carcinoma, epithelial–myoepithelial carcinoma, malignant myoepithelioma), and 6 cancers (2 MEC, 2 adenoid cystic carcinoma, CxPA, adenocarcinoma not otherwise specified) to adenomas and cyst [6, 7]. In the study presented by Godballe et al., in which 85 parotid gland carcinomas were reanalysed, diagnosis was changed after microscopic reassessment and immunohistochemical tests in 20 patients (23.5 %) [8]. In a large national study in Denmark, a revision of 886 cancers of the large and small salivary glands was done and diagnosis was changed in 121 of them (14 %). In 11 cases, the diagnosis was changed from carcinoma to adenoma, in 7 CxPA in situ was diagnosed, and in next 12 cases cancers appeared to be non-epithelial malignant tumours, and in 90 cases a subtype of cancer was changed. In one case, it was found that the cancer does not originate from the salivary glands [12]. Histological assessment of salivary gland neoplasms is difficult and requires specialist experience to avoid diagnostic traps, such as misdiagnosis of necrotizing sialometaplasia or squamous metaplasia within Warthin tumour as carcinoma [10, 12, 13]. Moreover, due to significant progress in the adjunct diagnostic procedures, diagnosis of parotid gland carcinomas may require immunohistochemical and molecular tests. Nowadays, many types of cancers of the salivary glands (adenoid cystic carcinoma, epithelial–myoepithelial carcinoma, MEC, MASC, hyalinizing clear cell carcinoma, CxPA, SDC, acinic cell carcinoma) have specific molecular biomarkers, which are used to confirm the diagnosis and also have prognostic significance [14]. Interesting results were presented by Bishop et al. After microscopic and immunohistochemical re-evaluation, and after applying molecular techniques (FISH) to acinic cell carcinoma (AcCC) specimens, the diagnosis was changed to an MASC in 9/11 (82 %) of tumours in intraoral location, 2/2 in submandibular gland, and only in 3 of 16 (19 %) in parotid [15]. Another problem faced by nonspecialist pathologists is lack of awareness of newly defined salivary neoplasms such as mucinous variant of myoepithelioma and MASC, as discovered in our study. This is illustrated in our study by the relatively high number of patients, whose diagnoses were changed from primary carcinoma of a salivary gland to a secondary lesion (metastasis to the parotid gland from kidneys, breast or skin). This is similar to the findings of Godballe et al., who reported revision of primary carcinomas to metastatic ones in 6 % of patients, with the primary location of the tumour in the breast, prostate, skin and lung [8]. Metastases to the parotid gland make 5–11 % of all malignancies of this gland with the vast majority of them originating in the skin on the head (squamous cell carcinoma and malignant melanoma) [16–18]. However, occasionally, the primary malignancy is located outside the head and neck (kidney, breast and lung), and the metastatic tumour can be its first symptoms [18, 19]. This illustrates why access to full clinical data is necessary for proper diagnosis [16–19].

Predicting the clinical course based on histology and progression of the disease is not obvious and to a great extent is subjective. In the studies by Van der Wal et al., during further follow-up of patients after histological reclassification there were no events observed to confirm the accuracy of diagnosis change [6, 7]. A change in the diagnosis from a malignant neoplasm to a benign one, a non-neoplastic lesion or an in situ cancer (CXPA), has a psychological significance for the patient; however, the practical (economic) aspect is important as well (shortening/conclusion of follow-up). Moreover, a change in diagnosis can occasionally enable new therapeutic options such as use of monoclonal antibody treatment (Trastuzumab, Cetuximab), kinases inhibitors BRAF, MTOR, MEK, androgen receptor blockers and others [14, 20].

Conclusions

Histological assessment of salivary gland carcinomas should be carried out by an experienced pathologist with an access to a specific panel of IHC and molecular tests. It is also crucial for pathologists to have access to patients’ full clinical data, especially the information about past treatment of other primary neoplasms. Histological reclassification/verification of parotid gland carcinomas can help explain the cause of atypical clinical course in some patients, and may sometimes enable clinicians to implement proper therapy at early stages of the disease.