Salivary gland small cell carcinoma is a rare neoplasm, accounting for less than 1% of salivary gland tumours. Little is known about the epidemiologic factors and treatment of this lesion. The authors report two cases and perform a systematic literature search from 1960 to 2011 for articles on salivary gland small cell carcinoma. Once the papers were reviewed, a database was generated to analyse clinical and pathological features, treatments and outcomes, and an attempt was made to identify prognostic factors. Available data were retrieved for 44 cases which fully satisfied the inclusion criteria; the median age was 64.25 years and the male:female ratio was 2.4:1. The parotid gland was the most common site (79.6%). The overall 1-, 2- and 5-year survival rates were 75.3%, 56.4% and 36.6%, respectively. According to the present review, patient age, tumour size, distant metastasis and cytokeratin-20 positive immunostaining were all significant prognostic factors in a univariate analysis. No particular treatment approach appeared to improve survival. This work reinforces knowledge about salivary gland small cell carcinoma’s epidemiologic features, and identifies new prognostic markers. The optimal management of this lesion remains controversial.
Small cell carcinoma (SmCC) was first described by Barnard in 1926 and was designated as oat cell sarcoma. Since then, SmCC has taken on various names and been found in multiples sites throughout the body, although it is most frequently diagnosed in the lungs. Salivary gland SmCC (SGSmCC) is extremely rare, accounting for less than 1% of all salivary gland tumours and fewer than 2% of salivary gland malignancies. This tumour has been described as involving both the major and minor salivary glands, with most cases affecting the parotid glands.
The diagnosis of SGSmCC has changed over time. In the past, some authors divided these lesions into neuroendocrine and non-neuroendocrine (ductal) types. A diagnosis of neuroendocrine SmCC was based on the presence of electron dense core granules and immunoreactivity for a neuroendocrine marker, while a diagnosis of the non-neuroendocrine type was based on the presence of ductal formations and the absence of neuroendocrine differentiation. Some authors have concluded that dividing SGSmCC into two histological subtypes is not justified, since these entities show no prognostic differences, and more recently the majority of cases have demonstrated neuroendocrine differentiation on immunohistochemistry analysis.
The newest AFIP classification establishes that SGSmCC are defined as a rare, primary malignant tumour composed of small, undifferentiated cells that exhibit neuroendocrine differentiation. This differentiation can be evidenced by immunohistochemical reactivity for at least one neuroendocrine marker such as chromogranin A, synaptophysin, CD57 (Leu-7), CD56 (neural cell adhesion molecule) and neurofilament. Immunoreactivity for neuron-specific enolase (NSE) alone is insufficient evidence to confirm neuroendocrine differentiation. Most SmCC are positive for cytokeratins, which often have a characteristic paranuclear dotlike pattern of reactivity, and 75% of cases are positive for cytokeratin 20 (CK20).
Some authors have recently suggested SGSmCC should be subdivided into cutaneous and pulmonary varieties, based on the presence or absence of CK20 expression. Although not fully accepted, this approach is based on the pulmonary variants of SGSmCC appearing to have a worse prognosis than the cutaneous variants.
SGSmCC exhibits a preference for males, with a peak incidence from the fifth to seventh decades of life. These patients typically present with a painless mass that develops rapidly over several months. A diagnosis of primary SGSmCC can only be confirmed after excluding the possibility of metastatic lesions.
Currently, 130 cases diagnosed as SGSmCC have been collected from the literature, consisting of data case reports and small series of cases. The rarity of SGSmCC, has meant that it has been difficult to establish treatment criteria or determine prognosis. The authors add two more cases of SGSmCC to the literature and present a systematic literature review to discuss the clinicopathological details, relevant criteria for diagnosis, biological behaviour, prognosis and treatment.
A systematic literature review was performed using electronic databases (PubMed, Medline and LILACS) to identify relevant publications between 1960 and 2011 that included cases of SGSmCC. The following search terms were used: ‘small cell carcinoma’, ‘small undifferentiated cell carcinoma’, ‘small anaplastic cell carcinoma’, ‘oat cell carcinoma’, ‘neuroendocrine carcinoma’ and ‘salivary gland’. Additionally, a manual search was conducted by cross-referencing from the retrieved manuscripts. Data on sociodemographic characteristics, clinical features, diagnostic criteria, treatment, follow up and outcomes were collected for each study. Studies lacking data, reporting inconsistent data regarding histological diagnosis, absence of immunopositivity for at least one neuroendocrine marker (excluding NSE), containing uncertainty regarding the salivary gland origin, or where retrieving specific data about the clinicopathologic features was difficult, were excluded. Cases where neuroendocrine differentiation was demonstrated only by electronic microscopy were not included. The authors also opted to exclude cases that originated in minor salivary glands, because of the difficulty in recognizing the lesion epicenter. For the prognostic analysis, only cases that clearly matched the criteria for SGSmCC morphological features and had immunohistochemical positivity for of at least one neuroendocrine marker, excluding NSE, were included and are presented in Table 1 .
|Case||Age/gender||Site||Pathological size (mm)||Metastasis||Treatment||Recurrences||Status/follow up (months)||Presence of neuroendocrine features||CK20 staining|
|1||69/M||SG||15||–||–||S||–||–||Leu-7: +/NSE +||–|
|3||80/F||PG||32||Yes||No||S + RND||Yes||DOD/13||Leu-7: +/Chromogranin + NSE +||–|
|4||67/F||SG||40||No||Yes||S + Rxt||Yes||DOD/51||Leu-7: +/Synaptophysin: +/NSE +||–|
|5||5/F||PG||25||No||No||Rxt + Cht||No||NED/72||Synaptophysin: Multifocally +||–|
|6||42/F||PG||20||No||No||S + Rxt||No||NED/75||Leu-7: +/NSE +||–|
|7||60/M||PG||25||No||No||S||No||DOC/53||Leu-7: +/NSE +||–|
|8||72/M||PG||35||No||Yes||S + Rxt||No||DOD/20||Leu-7: +/NSE +||–|
|9||65/M||PG||70||Yes||No||S + RND||No||NED/60||Chromogranin +/NSE +||–|
|10||63/M||PG||–||Yes||No||Rxt||No||NED/72||Chromogranin +/NSE +||–|
|11||76/M||SG||–||–||–||–||–||–||Leu-7: +/NSE +||–|
|12||82/F||PG||–||Yes||No||S + RND + Rxt||Yes||DOD/12||Leu-7: +/Synaptophysin: +/Chromogranin + NSE +||–|
|13||64/M||PG||–||No||No||S + RND + Rxt + Cht||No||NED/24||Synaptophysin +||–|
|14||31/M||PG||–||Yes||No||S + RND + Rxt + Cht||No||NED/6||Chomogranin: focal +/Synaptophysin +/Leu-7 difusely +||–|
|15||52/F||PG||55||Yes||No||S + RND + Rxt + Cht||No||NED/10||Chromogranin +/Leu-7+/NSE +||–|
|16||60/M||PG||60||No||No||S + RND + Rxt + Cht||No||NED/2||Chromogranin +/NSE +||–|
|17||78/M||PG||55||Yes||No||S + RND + Rxt||No||DOC/1||Synaptophysin +/NSE +||–|
|18||52/M||PG||–||Yes||No||S + RND + Rxt||Yes||DOD/12||Chromogranin +/NSE +||–|
|19||42/F||PG||–||No||No||S + Rxt||No||NED/18||Synaptophysin +/NSE +||–|
|20||81/M||PG||35||–||–||S||–||–||Synaptophysin +/NSE +||Positive (dot)|
|21||57/M||SG||–||Yes||No||S + RND + Rxt + Cht||No||NED/72||Synaptophysin +/Chromogranin +/NSE +||–|
|22||58/M||PG||25||No||Yes||S + Rxt||No||DOD/18||CD56 +/NSE: +||Negative|
|23||77/F||PG||18||No||No||S + RND + Rxt||No||NED/28||Chromogranin +/Synaptophysin +/Neurofilament +/CD56 +/Leu-7 +/NSE +||Positive (dot)|
|24||78/M||PG||15||Yes||Yes||S + RND + Rxt||No||DOD/45||Chromogranin +/Synaptophysin +/CD56(NCAM): +/Leu-7: +/NSE +||Positive (dot)|
|25||46/F||SG||100||Yes||No||S + RND + Rxt||Yes||DOD/6||Chromogranin +/Synaptophysin +/Neurofilament +/CD56 +/Leu-7 +/NSE +||Negative|
|26||47/M||PG||10||No||Yes||S||No||DOD/4||Chromogranin +/Synaptophysin +/NSE +||Negative|
|27||81/F||PG||30||Yes||Yes||S + RND||No||DOD/17||Synaptophysin +/NSE +||Positive|
|28||85/M||SG||38||No||Yes||S + RND||No||DOD/9||Chromogranin +/Synaptophysin +/NSE +||Positive (dot)|
|29||50/M||PG||7||Yes||No||S + RND||No||NED/155||Chromogranin +/Synaptophysin +/Neurofilament +/NSE +||Positive (dot)|
|30||70/M||PG||40||Yes||Yes||S + RND + Rxt||No||DOD/4||Chromogranin +/NSE +||Negative|
|31||66/M||PG||50||No||Yes||S + RND + Cht||No||DOD/20||Chromogranin +/Synaptophysin +/NSE +||Positive (dot)|
|32||76/M||PG||110||Yes||Yes||S + RND||No||DOD/2||Chromogranin +/NSE +||Positive (dot)|
|33||72/M||PG||29||No||No||S||No||NED/4||Chromogranin +/Synaptophysin +/Neurofilament +/CD56 +/Leu-7 +/NSE +||Positive (dot)|
|34||72/M||PG||20||Yes||Yes||S + RND + Rxt + Cht||No||DOD/34||Chromogranin +/Neurofilament +/NSE +||Positive (dot)|
|35||67/M||PG||80||Yes||Yes||S + RND + Rxt + Cht||No||AWD/18||Chromogranin +/NSE +||Positive (dot)|
|36||52/F||PG||15||No||No||S + RND + Rxt||No||NED/4||Chromogranin +/Synaptophysin +/Neurofilament +/NSE+||Positive|
|37||39/M||SG||25||No||No||S + Cht||No||NED/12||Chromogranin +/NSE +||–|
|38||74/F||PG||25||No||No||S + Rxt||No||NED/8||Chromogranin +/Synaptophysin +||Positive (dot)|
|39||46/F||SG||–||No||No||S + RND + Rxt + Cht||Yes||–||Chromogranin +/Synaptophysin +/NSE +||Negative|
|40||91/M||PG||–||Yes||No||Rxt||No||NED/36||Synaptophysin +/CD56 +||Positive (dot)|
|41||78/M||PG||–||Yes||Yes||S + RND + Rxt||Yes||DOD/36||Chromogranin +||Positive|
|42||77/M||PG||85||Yes||No||S + RND + Rxt + Cht||No||NED/10||Synaptophysin +/Leu-7 +||Positive (dot)|
|43 present||70/M||SG||50||No||Yes||Rxt + Cht||No||DOD/12||Chromogranin +(focal)/Leu-7 +/NSE +||Negative|
|44 present||69/M||PG||50||Yes||No||S||No||DOC/0||Chromogranin +/CD56 +/Leu-7 +/NSE +||Negative|
Only the overall mortality in survival reports was analysed. Overall survival curves for different clinicopathologic factors were constructed using the Kaplan–Meier method, and compared by the log-rank test and Cox proportional hazard model. A p value less than 0.05 was considered significant. The statistical analysis was conducted using Stata, version 10 (Stata Corp., College Station, USA). This study was reviewed and approved by the Institutional Committee for Ethics in Research (Protocol number: 149/11).
A 70-year-old white male was referred to the otolaryngology department complaining of a 4 month history of a rapidly enlarging painful mass in the right submandibular region. He was positive for HIV and HCV, and his past medical history included a basal cell carcinoma on the dorsum of his nose that was diagnosed two years previously (T3N0M0) and successfully treated with surgery and radiotherapy. A physical examination of the submandibular region revealed a firm, fixed mass approximately 5 cm in diameter. The thyroid was normal and no neck masses were identified by palpation and image exams. A computed tomography (CT) scan revealed a heterogeneous mass measuring 5 cm × 7 cm, with lobular contours and nonspecific borders, in the submandibular region ( Fig. 1 A ). The chest/abdomen radiograph and magnetic resonance imaging (MRI) examinations were normal, and no supposedly primary tumour could be detected. A fine-needle aspiration biopsy (FNAB) revealed malignant cells of unknown origin ( Fig. 1 B and C). The patient underwent an incisional biopsy that confirmed the diagnosis of T3N0M0 SGSmCC (Stage III). Because of the patient’s age and his clinical status, tumour remission was achieved using a combination of radiotherapy and chemotherapy. He received 5940 cGy over 33 fractions at a dose of 180 cGy per fraction to his right submandibular region. The chemotherapy consisted of three monthly cycles of etoposide plus cisplatin. During therapy, the patient was twice admitted with febrile neutropenia, and a progressive weight loss was noted. Five months after treatment, the patient was admitted to hospital with a diagnosis of icterus and cholestasis. An ultrasonogram of the abdomen revealed heterogeneous masses in the pancreas (8 cm 3 ), liver (2 cm 3 ) and one para-aortic lymph node that were consistent with metastatic disease. The patient was not considered for therapy because of the systemic metastases. One month later (12 months after diagnosis), the patient died of complications related to metastatic disease.
A 69-year-old white male was referred to the otolaryngology department complaining of a 12-month history of a painless mass in the right upper neck near the mandible, nodules in the neck region and a progressive unintentional weight loss. His past medical history was not significant. A physical examination of the neck revealed an approximately 4 cm firm mass on the right parotid and a 2 cm diameter enlarged ipsilateral cervical lymph node. There were no masses on the left side of the neck, and the thyroid was normal. An FNAB of the right neck nodule was performed, resulting in a diagnosis of T2N1M0 metastatic disease (Stage III). A complete work up that included X-ray examinations of the chest and abdomen proved negative in detecting any other lesion. A total parotidectomy with a radical right neck dissection was performed. One day after surgery, the patient developed respiratory distress, and an MRI of the chest showed an idiopathic pleural effusion (hemothorax). 2 days after surgery, the patient died from respiratory failure. The family refused a necropsy examination.
The two cases show similar cytological and histological patterns, and for this reason they are described together here. The Papanicolau-stained FNAB samples from both cases showed various clusters of small cells with scant cytoplasm and moderately pleomorphic round-to-ovoid nuclei with smooth or irregular contours containing dispersed thin, granular chromatin. No evidence of ductal formation or keratinization was found. The cells were dispersed among necrotic debris, erythrocytes and mononuclear inflammatory cells ( Fig. 1 B and C).
The hematoxylin–eosin stained histological slides from both cases showed malignant lesions formed by sheets, cords and irregular nests of small anaplastic cells containing nuclei that were two or three times larger than mature lymphocytes, were dispersed among dense fibrous stroma, and infiltrated the salivary gland tissue ( Fig. 2 A ). The tumour cells had scant cytoplasm and round-to-oval nuclei with absent or unremarkable nucleoli. Rosette-like and ductal structures, as well as squamous differentiation, were absent, while crush artifacts, mitotic figures and necrotic areas were found in the tumour ( Fig. 2 B). In both cases, focal areas of hemorrhage and discreet lymphocyte infiltration were noticed throughout the lesions.
Additionally in case 2, perineural invasion was clearly observed ( Fig. 2 C), and the surgical margins were compromised by the tumour. Small cells with more polygonal cytoplasm could also be seen ( Fig. 2 D). Neck dissection showed tumour in the submandibular gland and metastatic lesions in 19 lymph nodes.
Immunohistochemical evaluation of the tumours demonstrated positive immunoreactivity for epithelial and neuroendocrine markers. No positivity for lymphoid, neuronal or muscle markers was observed ( Table 2 ). In this way, the tumours were diagnosed as primary salivary gland small cell carcinomas.
|Antibodies to||Clone||Source||Dilution||Case I||Case II|
|Human epithelial antigen||BER-EP4||Abcam||1:100||+||NS|
|Epithelial membrane antigen||E29||Milipore||1:400||+||+|
|Thyroid transcription factor-1||8G7G3/1||Abcam||1:300||+||+|
|LCA (CD45)||PD7-26 e 2B11||Dako||1:100||−||−|
|Chromogranin A||DAK-A3||Dako||1:200||+ (focal)||+|