Mucoepidermoid carcinoma of salivary glands: A French Network of Rare Head and Neck Tumors (REFCOR) prospective study of 292 cases

Abstract

Background

To describe the characteristics of the largest European study of MEC of salivary glands and to determine the prognostic factors for overall and disease free survival.

Patients and methods

Patients with MEC were prospectively included in the Réseau d’Expertise Français sur les Cancers ORL Rares (REFCOR, French Network of Rare Head and Neck Tumors) database between 2009 and 2015.

Results

A total of 292 patients were included. Tumors were classified as low grade in 175 cases (60%), intermediate in 39 (13%) and high grade in 78 (27%). Median follow-up was 26 months. The 5-year OS and DFS rates were respectively 83% and 69%. In multivariate analysis, age (p = 0.004), diabetes (p = 0.02) and advanced stage (p = 0.03) were found to have a significant negative impact on OS. Diabetes (p = 0.001), alcohol consumption (p = 0.003) and advanced stage (p = 0.001) were found to have a significant negative impact on DFS. Compare to low grade, high grade tended to have a negative impact on OS (p = 0.05) and had a significant effect on DFS (0.002) while intermediate grade had no significant influence on survival. The surgical treatment had a positive impact on both OS (p = 0.00005) and DFS (p = 0.0005). Postoperative radiotherapy had no impact in multivariate analysis.

Conclusion

Advanced clinical stage, high grade tumor, high age, the impossibility of carrying out a complete surgical resection, and diabetes are the main prognostic factors in this prospective series of patients with MEC. Such findings open new research perspectives on the influence of these components on initial patient care.

Introduction

Salivary gland carcinomas represent less than 5% of malignant tumors of head and neck [1]. Parotid is the most common localization of these tumors among the major salivary glands, and palate is the most frequent site for minor salivary glands tumors. The prevalence is higher among women aged between 50 and 60 years.

Mucoepidermoid carcinoma (MEC) is the most common histology and accounts for approximately 30% of salivary gland carcinomas [2]. MEC are composed of squamous cells (epidermoid cells), mucinous cells, and a variable proportion of intermediate cells. These tumors are classified into three histological grades, from low to high grade correlated with their aggressiveness, according to the World Health Organization classification [3].

Complete surgical resection is the gold standard for MEC. The protocols for adjuvant radiotherapy or chemotherapy are still matter of discussion. The therapeutic strategy as well as the prognosis of the intermediate grade are still not well defined [4]. Several studies question the existence of the intermediate grades: the discovery of the MECT1-MAML2 translocation (mucoepidermoid carcinoma translocated1/Mastermidlike gene family) in molecular biology could favor the use of two histological grades instead of three [5].

Over the past ten years, the recent changes in clinical and histological classification criteria, the improvement of diagnostic imaging techniques and the advent of radiotherapy by IMRT have had an impact on the therapeutic strategy in MEC. The very low incidence of these cancers, makes it difficult to perform a randomized comparative trial.

The purpose of this article is to describe the clinical, histological and therapeutic features of the largest European study of MEC of salivary glands and to determine the prognostic factors for overall survival (OS) and disease free survival (DFS).

This study is a descriptive, prognostic, prospective, multicentric analysis. The inclusion of patients in the database was carried out by 35 centers of the Réseau d’Expertise Français des Cancers ORL Rares (REFCOR, French Rare Head and Neck Cancer Expert Network), using a standardized questionnaire. The data were anonymized, and an informed consent was signed by the patients in accordance with French law. Data quality control was ensured by file reviews. An update of the database was requested from the referring physicians. The inclusion period ran from January 2009 to December 2015. Ten patients, initially diagnosed between 1994 and 2009 and who presented with an oncological event requiring a new management between 2009 and 2015 were also included. Overall, 292 patients treated for a salivary gland MEC were included.

The inclusion criteria were: age >18 years, more than six months of follow-up and a defined histological grade. Sex, age of the patient at the time of diagnosis (in year), tobacco and alcohol consumption, longest time practiced profession according to the 24 categories established by the French official classification (INSEE [6]) and Karnofsky score were analyzed. Concerning therapeutic management, each case was discussed during the multidisciplinary tumor board of each center using REFCOR guidelines [32] and the experience of medical and surgical team. The tumors were described by the following criteria: site, histological grade (low (LG), intermediate (IG) or high (HG)), tumor size, nodal stage and metastatic disease. Histological analysis was performed according to the World Health Organization criteria and patients were clinically staged according to the American Joint Committee on Cancer (AJCC) TNM staging system and UICC 7th edition 2009 classification [7,8].

The characteristics of the patients and the tumors were described by the mean, median, minimum and maximum values for continuous data, and by their number and percentage for categorical data.

Overall survival (OS), disease-free survival (DFS) were estimated using the Kaplan-Meier method. The log-rank test was used to compare the survival functions. The multivariate analysis was performed using the Cox regression model, with a p-value = 0.10 as threshold for entering variable. Hazards Ratio (HR) and its 95% confidence interval (95% CI) were estimated. The variables age, sex, clinical stage, histological grade and resection margins were systematically tested in the multivariate analysis because of their established relevance in this pathology. The tests were performed bilaterally and were considered statistically significant when p ≤ 0.05. All the statistical analyses were performed on R® software (version 3.1.0).