Acta Otorrinolaringológica Gallega

Artículo Original

Tumores salivares malignos primários. Períodos diferentes, prognóstico semelhante?

Primary malignant salivary tumors. Different periods, similar prognosis?

João Carvalho de Almeida,

João Fernandes,

Eurico Monteiro

Porto Centre of Portuguese

Institute of Oncology

Correspondencia

João Carvalho de Almeida

Email:

joaocarvalhoalmeida@gmail.com

Fecha de envío: 21/12/2019

Fecha de aceptación: 1/5/2020

ISSN:
2340-3438

Edita:
Sociedad Gallega de Otorrinolaringología

Periodicidad:
continuada.

Web:
www.sgorl.org/ACTA

Correo electrónico:
actaorlgallega@gmail.com

Resumen

Introdução: Os tumores malignos das glândulas salivares são incomuns e representam aproximadamente 3-6% de todos os tumores malignos da cabeça e pescoço. A relativa raridade e a melhor compreensão da biologia destes tumores conduzem a progressos no diagnóstico, caracterização histológica e tratamento.

Métodos: Caraterizar e analisar o grupo de tumores malignos primários das glândulas salivares tratados no Instituto de Oncologia do Porto durante um período de 20 anos; agrupar os doentes de acordo com dois períodos diferentes (1995-2005) e (2006-2015) e comparar os resultados destes dois períodos.

Resultados: A média de idade foi de 60,88 anos. Não houve diferença de género. A glândula parótida foi a localização mais frequente, sendo responsável por 74,8% dos casos. A cirurgia foi o principal tratamento efetuado. A dissecção cervical foi realizada em 48,2% dos casos. A radioterapia adjuvante foi realizada em 79,5% dos doentes. A sobrevida de um ano e 5 anos foi de 0,800 (se=0,038) e 0,561 (se=0,048), respetivamente, e a sobrevida livre de doença de um ano e 5 anos foi de 0,701 (se=0,048) e 0,534 (se=0,054). Idade, adenopatias cervicais positivas e invasão perineural foram os fatores associados à sobrevida global. Os doentes tratados entre 1995 e 2005 apresentaram maiores taxas de recorrência e maior risco de morte, embora estas últimas diferenças não tenham sido estatisticamente significativas.

Conclusões: São necessários estudos epidemiológicos para compreender melhor a biologia e o comportamento dos tumores das glândulas salivares. Entretanto, a heterogeneidade e raridade desses tumores dificultam a realização de estudos prospetivos e randomizados.

Palabras clave glândulas salivares; oncologia; sobrevida livre de doença

Abstract

Background: Malignant salivary tumors are uncommon, and account for approximately 3-6% of all head and neck malignancies. The relative rarity and the better comprehension of the biology of these tumors conduced to progresses in the diagnosis, histological characterization and management.

Methods: Characterize and analyze the group of primary malignant tumors of salivary glands treated in the Porto Centre of Portuguese Institute of Oncology over a period of 20 years; group the patients according two different periods (1995 – 2005) and (2006 – 2015) and compare the results of these two periods.

Results: Mean age of the patients was 60.88 years. There was no gender difference. Parotid gland was the most frequent localization, being responsible for 74.8% of the cases. Surgery was the main treatment. Neck dissection was performed in 48.2% of the cases. Adjuvant radiotherapy was performed in 79.5% of the patients. One-year and 5-years survival were 0.800 (se=0.038) and 0.561 (se=0.048), respectively, and one-year and 5-years disease-free survival were 0.701 (se=0.048) and 0.534 (se=0.054). Age, positive cervical nodes and perineural invasion were the most important factors associated to overall survival. Patients treated between 1995 and 2005 had higher recurrence rates and a higher risk of death, although the latter differences were not statistically significant.

Conclusions: We consider that epidemiological studies are necessary to better understand the biology and behavior of salivary gland tumors. However, the heterogeneity and rarity of these tumors make hard to do prospective and randomized studies that could give us more information.

Keywords salivary glands; oncology; disease free survival

Introduction

Malignant salivary tumors are uncommon, and account for approximately 3-6% of all head and neck malignancies1. Global annual incidence of these tumors varies from 0.4 to 13.5 per 100.000 inhabitants.

Several histological tumors types may arise from major (parotid, submandibular and sublingual) or minor salivary glands.

The relative rarity and the better comprehension of the biology of these tumors conduced to progresses in the diagnosis, histological characterization and management2.

Surgery remains the cornerstone for the treatment of these tumors2 but complementary treatment options has been increasingly used in selected patients.

As in other head and neck tumors, better outcomes in salivary gland lesions have not improved significantly over time.

Nowadays, a detailed histological classification of these tumors, and the better characterization of prognostic factors may help in the future to predict clinical better outcomes and, identify most suitable treatment options.

Materials and methods:

All patients with primary malignant tumors of salivary glands presented for treatment in Porto Centre of Portuguese Institute of Oncology, between January 1995 and December 2015 were included.

Data collection was made consulting electronic medical records, in Glintt® application. Variables collected included gender, age, tumor location, histological classification, stage, treatment options, overall survival, disease-free survival, one and 5-years survival rates. Exclusion criteria included inadequate information, follow-up abandonment, and recurrent or distant disease at presentation.

Only definitive diagnoses were considered in the histopathological analysis and, only primary lesions of salivary glands were included, dismissing benign tumors, mesenchymal tumors, congenital anomalies, vascular and lymphatic malformations, granulomatous infections and salivary gland invasion by skin tumors.

All cases were recorded according to the 2005 World Health Organization (WHO) histological classification, published in 2005 and American Joint Committee on cancer TNM 7th edition.

Outcomes were defined in terms of overall survival and disease-free survival. One-year and 5-years survival rates were calculated.

Overall survival was defined as the time from diagnosis to the date of death. Disease free-survival was analyzed between the end of the first therapeutic intervention until the diagnosis of recurrent disease.

A database was built using IBM SPSS Statistics 24.0 for Mac operative system and, the same software was used for statistical analysis. Statistical significance level was defined as 5% (α = 0.05). Patients were characterized with simple description of the referred variables, related to the population characteristics and other parameters. Sample characteristics of categorical variables were presented as counts and percentages, while continuous variables were presented as means and standard deviation (sd) or median and interquartile range (IQR).

To analyze overall mortality, standard survival analysis methods were applied, considering death as the event of interest and time between discharge and death as the follow-up time. For patients alive at the end of the study, the follow-up time considered was time until the last contact.

To analyze disease-free survival, standard survival analysis methods were applied, considering recurrence as the event of interest and time between discharge and recurrence as the follow-up time. Kaplan-Meier curves were obtained for total sample in both analyses. Cox proportional hazards regression models were used to analyze the effect of covariates on the overall mortality after discharge and on the disease-free survival.

In both analyses, to decide which variables should be included in the multivariable model, an exploratory analysis was performed by fitting univariable models and considering as candidates for the multivariable models all variables significant at the 0.05 significance level in these univariable models. For the final multivariable models, the significance level was built including all variables with p-values lower than 0.05.

The data of our study was divided into two periods, period 1 (1995–2005) and period 2 (2006–2015). The purpose was to evaluate changes in patient’s characteristics and outcomes.

Both groups were characterized with simple description of the variables, related to gender, age, tumor location, histological diagnosis, stage, treatment options, adjuvant therapy, one and 5-years survival, overall survival, disease-free survival and recurrence. These variables were compared in the two groups using independent samples t test or Mann-Whitney test for continuous and Chi-square test (with continuity correction in the case of 2x2 table) or Fisher’s exact test for categorical variables.

Results

Sample characteristics

The characteristics of the 139 patients included in the study are represented in Table 1. Fifty eight patients (41.7%) belong to the period 1 (1995-2005) and 81 (58.3%) to the period 2 (2006-2015). The median follow-up in the two groups was 5.4 years (64.8 months).

Patients were equally distributed in terms of gender, (73 males and 66 females). Average age was 60.88 years (sd - 16.35). Mean age at onset was higher in females [58.39 years (sd - 17.79)] than in males [63.12 years (sd - 14.70)].

The most common tumor localization was the parotid gland (74.8%), followed by submandibular gland (18.7%), minor salivary glands (5.8%) and sublingual gland (0.7%) – Figure 1.

The most frequent histological type was mucoepidermoid carcinoma (18%), followed by squamous cell carcinoma (17.3%) and adenoid cystic carcinoma (16.5%) – Figure 2.

Tumors distribution by site is showed in Table 2.

Among the 25 mucoepidermoid carcinomas, 84% were in the parotid gland, 12% in the submandibular gland and 4% in minor salivary glands.

Regarding squamous cell carcinomas, parotid gland was also the most common localization, with 83.3% of the cases.

Among the 23 cases of adenoid cystic carcinomas, 47.8% were in the parotid gland, followed by the submandibular gland (34.8%), minor salivary glands (13%) and sublingual glands (4.3%).

Treatment options and post-operative parameters are represented in Table 3.

In more than half of the patients the disease presented in early-stage (stage I 24.5%, stage II 38.1%, stage III 18%, stage IV 19.4%) at diagnosis – Table 3 and Figure 3.

99.3% of patients in our sample were treated with surgery as primary option. Only one patient was not submitted to surgery, receiving radiotherapy in combination with chemotherapy because the tumor was considered non-resectable.

Free margins were found in 36.7% of the patients. Lymph node dissection was performed in 48.2% of the cases and positive lymph nodes metastasis in histopathological examination were found in 28.2% of the cases (mainly from patients with squamous cell carcinoma, salivary duct carcinoma and adenocarcinoma – 11.7%, 4.9% and 2.9% - respectively).

Adjuvant radiation therapy was administered in 79.5% of patients, namely in those with positive resection margins, regional metastasis, high-grade histology’s or in the presence of aggressive histologic features (cytologic atypia, increased mitotic activity and necrosis). 8.9% of the patients received adjuvant chemotherapy. The association of chemotherapy occurred as adjuvant and in combination with radiotherapy. Recurrence was observed in 49 patients (48.5%).

Comparison between the two periods

According to one of the aims of the study, patients were divided in two periods and different variables were compared in order to identify and contrast several features – Table 1.

In the period 1 were included 31 female (53.4%) and 27 male patients (46.6%). Average age in this group was 57.16 years (sd - 18.16). Concerning the primary site, 81% of the tumors affected the parotid gland, followed by submandibular gland (15.5%) and minor salivary glands (3.4%). The most common histology was adenoid cystic carcinoma (19%), followed by mucoepidermoid carcinoma (17.2%), squamous cell carcinoma (15.5%), acinic cell carcinoma (13.8%), adenocarcinoma (10.3%) and salivary duct carcinoma (10.3%). Remaining histological types were responsible by less than 10% of the cases.

In the period 2 were included 35 female (43.2%) and 46 male patients (56.8%). Average age in this group was 63.54 (sd - 14.46). Concerning the primary site, 70.4% of the tumors were located in the parotid gland, followed by submandibular gland (21%), minor salivary glands (7.4%) and sublingual gland (1.2%). The most common histologies were squamous cell carcinoma and mucoepidermoid carcinoma (18.5% each), salivary duct carcinoma (15.5%) and adenoid cystic carcinoma (14.8%). Remaining histological types were responsible by less than 10% of the cases.

As observed in Table 1, patients from the period 1 were on average 6 years younger than patients of the period 2 with differences statistically significant [t(105) = 2.222, p = 0.028].

Regarding stage, in the period 1 more than half of the patients presented in early-stage disease (stage I 22.4%, stage II 37.9%, stage III 19%, stage IV 20.7%). In the period 2, more than half of the patients presented also in early-stage disease (stage I 25.9%, stage II 38.3%, stage III 17.3%, stage IV 18.4%) – Table 2 and Figure 3.

Speaking about treatment options surgery was performed in 98.3% of the patients in the period. Only one patient was not submitted to surgery, effectuating radiotherapy in combination with chemotherapy. Has other findings, free margins were obtained in 51.1% of the patients. Positive resection margins were present in 48.9% of the patients. Lymph node dissection was performed in 48.6% of the cases. Positive lymph nodes in pathological examination were found in 33.3% of the cases. In this group, 23.1% of the cases had perineural invasion (patients with adenoid cystic carcinoma, carcinoma ex pleomorphic adenoma, adenocarcinoma, acinic cell carcinoma, oncocytic carcinoma and clear cell carcinoma – 3.8% each). 50% were high or intermediate grade (mostly mucoepidermoid carcinoma, adenoid cystic carcinoma and salivary duct carcinoma with 13.2% the first two and 10.6% the latter). Adjuvant radiation was effectuated in 86.8% of the patients. 9.4% of these patients received adjuvant chemotherapy. Recurrence was observed in 28 patients (60.9%).

In the period 2, surgery was performed in 100% of patients. Free margins were found in 39.7% of cases and 60.3% had positive resection margins. Lymph node dissection was performed in 47.9% of patients and positive lymph nodes were observed in 25.7%. On histological examination 45.3% of the tumors had perineural invasion (mainly adenoid cystic carcinoma, squamous cell carcinoma and salivary duct carcinoma – 14%, 11% and 6.3% respectively). 72.9% were high or intermediate grade (adenoid cystic carcinoma in 25% of cases and 14.6% in mucoepidermoid and salivary duct carcinomas). Adjuvant radiation was administered in 73.9% of patients in this group and in 8.6% adjuvant chemotherapy. Recurrence was observed in 21 patients (28.2%).

Looking for all the variables, both groups demonstrate similar results regarding most of the parameters analyzed as observed in Table 2.

However, it can be noted that patients treated in the period 1 (1995-2005) had a higher recurrence rate (period 1 60.9% vs. period 2 28.2%, p = 0.038).

Survival analysis – overall survival

A hundred ten patients were included for overall survival analysis.

Figure 4 represents Kaplan-Meyer curves for overall survival in the total sample. One-year and 5-years survival rates were 0.800 (se=0.038) and 0.561 (se=0.048) respectively, and the median was 74.0 months (95% CI 34.98-113.0).

Table 4 includes information about unadjusted and adjusted hazard ratio (HR) and 95% confidence interval (95% CI) for overall mortality.

For the variable primary site, we consider in the survival analysis, two categories: 1) parotid vs 2) submandibular + sublingual + minor salivary glands, and for the variable stage were considered also two categories: 1) stage I and stage II; 2) stage III and stage IV.

As we can observe in Table 4, variables as age, stage, grade, lymph node dissection, pN+, surgical margins and perineural invasion, were identified as potential predictive factors for overall mortality with p values < 0.05 on Cox-model.

After adjusting for gender and age (Table 4), only age, pN+ and perineural invasion remain statistically significant on statistical analyses.

Risk of death is higher in older (HR=1.053, 95% CI 1.014-1.094) and pN+ patients (HR=3.857, 95% CI 1.200-12.40) and in the group with perineural invasion (HR=3.175, 95% CI 1.188-8.484).

Survival analysis – disease-free survival (recurrence)

For analysis of disease-free survival, a total of 135 patients were included.

Figure 5 presents Kaplan-Meyer curve for disease-free survival in all sample. One-year and 5-years disease-free survival estimate were 0.701 (se=0.048) and 0.534 (se=0.054), respectively, and the median was 72.0 months (95% CI 20.28-123.7).

Table 5 includes information about unadjusted and adjusted hazard ratio (HR) and 95% confidence interval (95% CI) for disease-free survival.

For the variable primary site associated to survival analysis, we consider the same two categories: 1) parotid vs 2) submandibular + minor salivary glands + sublingual. For the variable stage, two categories were also considered: 1) stage I and stage II; 2) stage III and stage IV.

As demonstrated in Table 5, variables such as lymph nodes dissection, pN+, grade, surgical margins and perineural invasion and stage were identified as potential predictive factors of disease-free survival, with p values < 0.05 on Cox-model.

After adjusting for gender and age, as we can observe in Table 5, no one of the variable remains statistically significant.

Survival analysis – comparison between the two periods

Regarding overall survival and disease-free survival, Figure 6 and 7 represent Kaplan-Meyer curves comparing the two periods. One-year and 5-years overall survival rates were 0.720 and 0.458 for period 1 and 0.867 and 0.644 for period 2. Regarding one-year and 5-years disease free-survival, values were 0.605 and 0.414 for the period 1 and 0.770 and 0.630 for the period 2.

Discussion

This study analyzes in several contexts patients with different salivary gland tumors treated in two selected periods in the Porto Centre of Portuguese Institute of Oncology.

Mean age of the patients was 60.88 years, similar to the referred in the literature. A recent study pointed to a mean age of 55.66 years3. Other articles report mean ages ranging from 45.2 years to 58.6 years4. The higher mean age of the patients in our study could be related with a high percentage of squamous cell carcinomas, that some studies report as appearing in advanced ages at diagnosis3.

There was no gender difference in the patients of our study, with a male:female ratio of 1.1, similar to the referred by Luksic et al5. However, some other authors have found salivary gland tumors to be more frequent in females4,6.

Parotid gland was the most frequent localization, being responsible for 74.8% of the cases, aspect also observed in most large case series4,6–9. Submandibular gland was the second most frequent localization with 18.7% of the cases, also similarly to other reports10.

The frequency of histological tumor types ranges from series to series.

In contrast to most head and neck tumors, in which squamous cell histology predominate, salivary gland tumors encompass distinct and varied histological subtypes11. In our series, mucoepidermoid carcinoma was the most frequent histopathological diagnosis (18%), aspect also referred in several other series4,8 , followed by squamous cell carcinoma (17.3%) and adenoid cystic carcinoma (16.5%). Percentages of mucoepidermoid carcinoma and adenoid cystic carcinoma in some consulted series were similar to our study6. However, it’s not usual the reference to so high percentage of squamous cell carcinoma histologies11,12. The explanation for this difference could be in relation to misclassification, namely the categorization of high-grade mucoepidermoid carcinomas as squamous cell carcinomas.

In regard to adenoid cystic carcinoma, it is known that it frequently affects submandibular gland and minor salivary glands13, aspect also observed in the present analyzes.

Speaking about treatment options, surgery with or without neck dissection was the main treatment in our patient’s series, the same option than the referred in other analysis. Generally, patients with poor overall health status and presence of comorbidities do not perform surgery. In our study, only 1 patient has not been submitted to surgery because the tumor was non-resectable.

In respect to neck dissection, which was performed in 48.2% of the cases, the type of dissection was selected in accordance with the extent of nodal disease in cases of clinically or imagiology positive neck. Adjuvant radiotherapy, in turn, was performed in high grade tumors, positive neck, stage III/IV, surgical positive margins, perineural invasion and high-risk histological types, totaling 79.5% of the patients.

One-year and 5-years survival were 80.0% and 56.1%, respectively, and one-year and 5-years disease-free survival were 70.1% and 53.4%, which is in accordance with the referred in literature. A recent study of 2016, analyzing malignant salivary gland tumors in a specific Spanish Mediterranean population and including all tumors between January 2004 and December 2012, report a 5-year overall survival rate of 52.5% and 5-year disease-free survival rate of 50%11, which is in accordance with our results.

In regard to overall survival associated to known and accepted prognostic factors, as summarized in Table 4, we demonstrate that only age, positive nodes and perineural invasion were statistically significant. In general, these parameters are considered to be closely related to decreased survival and could predict the risk of death11,14. Perineural invasion, particularly, has been considered in several studies as an important predictor of more aggressive tumor behavior15,16. Our findings are in accordance with these results, and perineural invasion in our patients population was one of the factors associated to poorer rates of overall survival.

Regarding disease free-survival as summarized in Table 5, none of the referred predictive factors were statistically significant, namely age, positive nodes or perineural invasion. None of these variables could predict the risk of recurrence. This could be in relation with some of the limitations of our study, namely the lack of data, due to the length of time analyzed and the heterogeneity of pathologists. Even consulting the literature, a high variability in identifying predictive factors for disease recurrence it’s observed. Available studies are retrospective and prospective studies analyzing potentially prognostic factors for disease recurrence are scarce.

As mentioned before, according to one of the purposes of this study, patients were divided in two periods and were compared concerning different variables. In this context, regarding age, we can observe that patients from the period 1 were, on average, 6 years younger than patients of period 2 and these differences were statistically significant. Consulting the literature, several studies refer that the age of onset of the disease is lower in females. In one study from Spain, for instance, mean age at onset in females was 52.92 years and 58.43 years in males3. In fact, in our study, patients from the period 1 had a higher percentage of females (53.5% vs. 43.2%), and this fact could explain the different age of onset in the two groups.

Concerning other parameters, this study revealed differences between the two periods according recurrence rate. Patients treated in the period 1 (1995-2005) had higher recurrence rates (60.9% vs. 28.2%), with differences statistically significant (<0.05). As demonstrated in our disease free-survival analysis, none of the predictive factors accepted for recurrence as age, positive nodes or perineural invasion, could predict in our patients increased risk for recurrence. As a possible explanation, advances in radiation therapy protocols, with more precise targets for radiation, may improve local control rates.

As showed before, patients treated in the period 1 had a higher risk of death, notwithstanding with differences statistically not significant (p > 0.05). Again, this could be in relation with the same justifications above.

This study has several limitations. First of all it’s a retrospective analysis and has lack of data due to the length of time analyzed. This leads us to believe that with more detailed data, it could be possible to ascertain more precisely the association between different variables. Secondly, this study has the inherent heterogeneity of surgical techniques, anatomopathological details and treatment options, namely in the field of radiotherapy.

Another shortcoming is the sample size that is small and some recent identified histological subtypes were not identified.

Further epidemiological studies are necessary to improve the comprehension of salivary gland tumors course, and learn to develop strategies either in diagnosis as in the management. However, the heterogeneity and rarity of this type of tumors make hard to do prospective and randomized trials that could give us much more information.

Conclusions

The distribution of salivary gland tumors in our analysis is similar to several other reports from different countries.

Our findings indicate that age, positive cervical nodes and perineural invasion are the most important factors associated to overall survival.

Analyzing two temporal periods we observe that patients treated between 1995 and 2005 (period 1) had higher recurrence rates and a higher risk of death, in comparison with patients of the period 2 although the latter differences were not statistically significant (p > 0.05).

Concluding, we consider that epidemiological studies are necessary to better understand the biology and behavior of salivary gland tumors. However, the heterogeneity and rarity of these tumors make hard to do prospective and randomized studies that could give us more information.

Declaración de conflicto de intereses

The authors declare that they have no conflict of interest with respect to this article.

Bibliography

1. Bradley PJ. Classification of Salivary Gland Neoplasms. 2016;78:1-8. doi:10.1159/000442119

2. Jegadeesh N, Liu Y, Prabhu RS, et al. Outcomes and prognostic factors in modern era management of major salivary gland cancer. ORAL Oncol. 2015. doi:10.1016/j.oraloncology.2015.05.005

3. Collazo-fernández L, Campo-trapero J, Cano-sánchez J, García-martín R. Retrospective study of 149 cases of salivary gland carcinoma in a Spanish hospital population. 2017;22(2). doi:10.4317/medoral.21419

4. Bell RB, Dierks EJ. Management and Outcome of Patients With Malignant Salivary Gland Tumors. 2005:917-928. doi:10.1016/j.joms.2005.03.006

5. Luk I. Salivary gland tumours : 25 years of experience from a single institution in Croatia. 2012;40. doi:10.1016/j.jcms.2011.05.002

6. Long-jiang L, Yi L, Yu-ming W, Hua L, Hong-wei Z. Clinical analysis of salivary gland tumor cases in West China in past 50 years. 2008:187-192. doi:10.1016/j.oraloncology.2007.01.016

7. Dilci A, Aydil U, Du M, Ko A. Salivary Gland Tumors in Turkey : Demographic Features and Histopathological Distribution of 510 Patients. 2013;65(July):112-120. doi:10.1007/s12070-012-0594-6

8. Wang X, Meng L, Hou T, Huang S. Tumours of the salivary glands in northeastern China : a retrospective study of 2508 patients. Br J Oral Maxillofac Surg. 2014. doi:10.1016/j.bjoms.2014.10.008

9. Kara Mİ, Göze F, Ezirganlı Ş, Polat S, Muderris S, Elagoz S. Neoplasms of the salivary glands in a Turkish adult population. 2010;15(6):6-11. doi:10.4317/medoral.15.e880

10. Saghravanian N, Ghazi N, Saba M. Annals of Diagnostic Pathology Clinicopathologic evaluation of salivary gland neoplasms : a 38-year retrospective study in Iran. Ann Diagn Pathol. 2013;17(6):522-525. doi:10.1016/j.anndiagpath.2013.05.008

11. Ata-ali J, Zurriaga O, Alberich C. Incidence and survival rates for malignant salivary gland tumors. 2016;(March). doi:10.2334/josnusd.58.67

12. Boukheris H, Curtis RE, Land CE, et al. Incidence of Carcinoma of the Major Salivary Glands According to the WHO Classification , 1992 to 2006 : A Population-Based Study in the United States Incidence of Carcinoma of the Major Salivary Glands According to the WHO Classification , 1992 to 2006 : A Population-Based Study in the United States. 2009. doi:10.1158/1055-9965.EPI-09-0638

13. Ito FA, Ito K, Vargas PA, Almeida OP De, Salivary MAL. Salivary gland tumors in a Brazilian population : a retrospective study of 496 cases. 2005:533-536. doi:10.1016/j.ijom.2005.02.005

14. Israel Y, Rachmiel A, Gourevich K, Nagler R. Mortality rates and prognostic factors in patients with malignant salivary tumors. Med Oncol. 2019;36(7):1-9. doi:10.1007/s12032-019-1284-y

15. Eijkenboom WMH, Marres HAM. Salivary gland carcinoma : independent prognostic factors for locoregional control , distant metastases , and overall survival : results of the dutch head and neck oncology cooperative group. 2004;(August):681-693. doi:10.1002/hed.10400

16. Erhaard CHHJT, Ubsen HEL, Aanders MK, et al. Head and Neck.The role of radiotherapy in the treatment of malignant salivary gland tumors p iet l . a . van d en e nde , m . d ., # f red b urlage , m . d .,** on behalf of the d utch. 2005;61(1):103-111. doi:10.1016/j.ijrobp.2004.03.018

Tables y Figures

Table 1 – Patient characteristics including histology.

	Total (n=139)	1995-2005 (n=58; 41.7%)	2006-2015 (n=81; 58.3%)	Statistical test	p
Gender Female, n, (%) Male, n, (%)	66 (47.5) 73 (52.5)	31 (53.4) 27 (46.6)	35 (43.2) 46 (56.8)	χ2(1)=1.040	0.308
Mean age, years , (sd)	60.88 (±16.35)	57.16 (±18.16)	63.54 (±14.46)	t(105)=2.222	0.028
Primary Site Parotid, n (%) Submandibular, n (%) Minor Salivary Glands, n (%) Sublingual, n (%) Histology Squamous Cell Carcinoma, n (%) Mucoepidermoid Carcinoma, n (%) Adenoid Cystic Carcinoma, n (%) Carcinoma Ex Pleomorphic Adenoma, n (%) Adenocarcinoma, NOS, n (%) Acinic Cell Carcinoma, n (%) Salivary Duct Carcinoma, n (%) Lymphoma, n (%) Oncocytic Carcinoma, n (%) Myoephitelial Carcinoma, n (%) Clear Cell Carcinoma, NOS, n (%) Cystadenocarcinoma, n (%) Carcinosarcoma, n (%)	104 (74.8) 26 (18.7) 8 (5.8) 1 (0.7) 24 (17.3) 25 (18) 23 (16.5) 4 (2.9) 6 (4.3) 14 (10.1) 15 (10.8) 6 (4.3) 4 (2.9) 9 (6.5) 3 (2.2) 1 (0.7) 2 (1.4) 3 (2.2)	47 (81) 9 (15.5) 2 (3.4) 0 (0) 9 (15.5) 10 (17.2) 11 (19) 2 (3.4) 6 (10.3) 8 (13.8) 6 (10.3) 1 (1.7) 1 (1.7) 1 (1.7) 2 (3.4) 0 (0) 1 (1.7) 0 (0)	57 (70.4) 17 (21) 6 (7.4) 1 (1.2) 15 (18.5) 15 (18.5) 12 (14.8) 2 (2.5) 0 (0) 6 (7.4) 9 (15.5) 5 (6.2) 3 (3.7) 8 (9.9) 1 (1.2) 1 (1.2) 1 (1.2) 3 (3.7)	- -	- -

Table 2 – Distribution of tumors by site.

	Parotid Gland (n=104; 74.8%)	Submandibular Gland (n=26; 18.7%)	Minor Salivary Glands (n=8; 5.8%)	Sublingual Glands (n=1; 0.7%)	Total (n=139)
Histology, n (%) Squamous Cell Carcinoma Mucoepidermoid Carcinoma Adenoid Cystic Carcinoma Carcinoma Ex Pleomorphic Adenoma Adenocarcinoma, NOS Acinic Cell Carcinoma Salivary Duct Carcinoma Lymphoma Oncocytic Carcinoma Mioephitelial Carcinoma Clear Cell Carcinoma, NOS Cystadenocarcinoma Carcinosarcoma Basal Cell Adenocarcinoma	20 (19.2) 21 (20.2) 11 (10.6) 4 (3.8) 6 (5.8) 11 (5.6) 13 (12.5) 3 (2.9) 2 (1.9) 8 (7.7) 1 (0.9) 1 (0.9) 1 (0.9) 2 (1.9)	2 (7.7) 3 (11.5) 8 (30.8) 0 (0) 0 (0) 2 (7.7) 2 (7.7) 3 (11.5) 2 (7.7) 0 (0) 2 (7.7) 0 (0) 1 (3.8) 1 (3.8)	2 (25) 1 (12.5) 3 (37.5) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 1 (12.5) 0 (0) 0 (0) 0 (0) 0 (0)	0 (0) 0 (0) 1 (100) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0)	24 25 23 4 6 14 15 6 4 9 3 1 2 3

Table 3 – Treatment options and other tumor parameters.

	Total (n=139)	1995-2005 (n=58; 41.7%)	2006-2015 (n=81; 58.3%)	Statistical test	p
Surgery (primary site), n (%)	138 (99.3)	57 (98.3)	81 (100)	-	-
Lymph Node Dissection, n (%) Yes, n (%) No, n (%) pN+, n (%) Yes, n (%) No, n (%)	53 (48.2) 57 (51.8) 29 (28.2) 74 (71.8)	18 (48.6) 19 (51.4) 11 (33.3) 22 (66.7)	35 (47.9) 38 (52.1) 18 (25.7) 52 (74.3)	χ 2(1)=0.000 χ 2(1)=0.322	1.000 0.570
P Stage				-	-
I, n (%) II, n (%) III, n (%) IVa, n (%) IVb, n (%) IVc, n (%) Grade High / Intermediate Grade, n (%) Low Grade, n (%) Surgical Margins Positive, n (%) Negative, n (%) Perineural Invasion Yes, n (%) No, n (%) Adjuvant Radiation, n (%) Yes, n (%) No, n (%) Adjuvant Radiation/Chemotherapy, n (%) Recurrence, n (%) Yes, n (%) No, n (%)	34 (24.5) 53 (38.1) 25 (18) 22 (15.8) 4 (2.9) 1 (0.7) 54 (62.8) 32 (37.2) 64 (55.7) 51 (44.3) 35 (8.9) 55 (61.1) 97 (79.5) 25 (20.5) 8 (8.9) 49 (48.5) 52 (51.5)	13 (22.4) 22 (37.9) 11 (19) 9 (15.5) 3 (5.2) 0 (0) 19 (50) 19 (50) 23 (48.9) 24 (51.1) 6 (23.1) 20 (76.9) 46 (86.8) 7 (13.2) 3 (9.4) 28 (60.9) 18 (39.1)	21 (25.9) 31 (38.3) 14 (17.3) 13 (16) 1 (1.2) 1 (1.2) 35 (72.9) 13 (27.1) 41 (60.3) 27 (39.7) 29 (45.3) 35 (54.7) 51 (73.9) 18 (26.1) 5 (8.6) 21 (28.2) 34 (61.8)	χ 2(1)=3.837 χ 2(1)=1.029 χ 2(1)=2.968 χ 2(1)=2.313 - χ 2(1)=4.294	0.050 0.310 0.085 0.128 - 0.038

Table 4 – Univariable and multivariable Cox-model for overall mortality after discharge.

	Unadjusted			Adjusted
	HR	95% CI	p	HR	95% CI	p
Gender
Male	1.332	0.802-2.210	0.268	1.836	0.757-4.456	0.179
Female	1	-	-	1	-	-
Age	1.021	1.002-1.039	0.026	1.053	1.014-1.094	0.007
Period
1995-2005	1.641	0.972-2.773	0.064
2006-2015	1	-	-
Primary site
Parotid	1.747	0.928-3.289	0.084
Other	1	-	-
Lymph Node Dissection
Yes	3.155	1.616-6.159	0.001	0.606	0.194-1.897	0.390
No	1	-	-	1	-	-
pN+
Yes	4.672	2.352-9.281	<0.001	3.857	1.200-12.40	0.023
No	1	-	-	1	-	-
Stage
I and II	1	-	-	1	-	-
III and IV	2.286	1.370-3.814	0.002	1.824	0.703-4.737	0.217
Grade
High/intermediate grade	2.702	1.211-6.031	0.015
Low grade	1	-	-
Surgical margins
Positive	3.130	1.606-6.103	0.001	0.438	0.150-1.278	0.131
Negative	1	-	-	1	-	-
Perineural invasion
Yes	2.283	1.097-4.749	0.027	3.175	1.188-8.484	0.021
No	1	-	-	1	-	-
Adjuvant radiation
Yes	0.790	0.416-1.500	0.472
No	1	-	-

Table 5 – Univariable and multivariable Cox-model for disease-free survival.

	Unadjusted			Adjusted
	HR	95% CI	p	HR	95% CI	p
Gender
Male	1.306	0.733-2.324	0.365	1.291	0.515-3.237	0.585
Female	1	-	-	1	-	-
Age	1.004	0.986-1.023	0.660	1.035	0.998-1.073	0.061
Period
1995-2005	1.719	0.958-3.088	0.070
2006-2015	1	-	-
Primary site
Parotid	1.571	0.796-3.102	0.193
Other	1	-	-
Lymph Node Dissection
Yes	2.622	1.333-5.159	0.005	0.682	0.215-2.161	0.515
No	1	-	-	1	-	-
pN+
Yes	3.075	1.423-6.645	0.004	1.843	0.534-6.362	0.333
No	1	-	-	1	-	-
Stage
I and II	1	-	-	1	-	-
III and IV	2.345	1.314-4.185	0.004	2.260	0.764-6.690	0.141
Grade
High/intermediate grade	2.920	1.161-7.345	0.023
Low grade	1	-	-
Surgical margins
Positive	3.508	1.702-7.231	0.001	0.987	0.351-2.780	0.981
Negative	1	-	-	1	-	-
Perineural invasion
Yes	3.394	1.550-7.434	0.002	2.299	0.892-5.921	0.085
No	1	-	-	1	-	-
Adjuvant radiation
Yes	1.003	0.484-2.078	0.994
No	1	-	-

Figure 1 – Tumor localization.

Figure 2 – Histological type.

Figure 3 – Stage.

Figure 4 – Kaplan-Meier curve for overall mortality.

Figure 5 – Kaplan-Meier curve for disease-free survival.

Figure 6 – Kaplan-Meier curve comparing overall survival between the 2 periods.

Figure 7 – Kaplan-Meier curve comparing disease-free survival between the 2 periods.