Clinical and Prognostic Factors of Submandibular Gland Malignancies: A SEER-Based Analysis

Salivary gland malignancies encompass approximately 3–5% of head and neck cancers with 5–15% isolated from the submandibular gland (SMG) [1‐4]. The preponderance of major salivary gland malignancies has steadily increased. Most are found in the parotid and submandibular glands and fewer cases reported within sublingual and minor salivary glands [1]. Relative to the parotid, SMG neoplasms have a rate of malignancy nearly twice as high and are associated with poorer prognoses [2].

Numerous exposures have been implicated in the development of salivary gland neoplasms, including modifiable factors such as alcohol and tobacco usage. Benign and malignant SMG neoplasms encompass a spectrum of histological types, and in most cases, surgical removal of the submandibular gland with possible adjuvant radiation therapy is the gold standard [3]. A neck dissection can be performed at the time of the initial surgery depending on the histology and stage. Other treatments in the literature include radiotherapy, chemoradiotherapy, and chemotherapy. Various factors contribute to survival including extent of lymph node involvement, local tissue invasion, and distant metastasis. However, information about these tumors' clinical and prognostic factors remains limited due to their rarity.

In recent years, few updates have been made synthesizing the epidemiology and prognostic factors of SMG neoplasms and various treatment paradigms [1‐5]. We aim to analyze the Surveillance, Epidemiology, and End Results Program (SEER) database to assess outcomes of SMG malignancy patients and reveal any relationships between clinical or prognostic factors [6].

The submandibular gland accounts for a small fraction of salivary gland tumors. Previous studies estimate approximately half of SMG tumors as benign and half as malignant [8]. Due to their low occurrence rates, the current literature is limited on epidemiology, rates of histological subtypes, and survival. The SEER database lends a greater sample size with data that has been employed to examine survival for other malignancies. Multi-variate analysis was also able to find multiple independent correlations with overall survival including demographics, histological subtype, grade, stage at presentation, size, and treatment modality. Another study conducted by Lee et al. examined survival for SMG tumors using SEER data from 1973 to 2011 and discussed correlates with size, stage, and type [2]. Our study updates and builds upon this work by assessing how survival is impacted by treatment modality for different histological subtypes and tumor grades.

The six reported subtypes in our analysis comprised over 94% of sample within the SEER database. The additional “Other” category included some other rarer subtypes such as acinar, mucinous/serous, and sarcomas/soft tissue tumors. Adenocarcinomas made up the majority of reported SMG tumors, though we suspect that this is including adenoid cystic carcinoma which is the most common histology reported in the literature, followed by squamous cell carcinoma and mucoepidermoid [10]. Previous studies similarly show a preponderance of adenoid cystic carcinoma, squamous cell carcinoma, and mucoepidermoid carcinoma, with rates in the literature reporting 36%, 18.1%, and 16.9%, respectively [2, 9, 11]. Notably, primary squamous cell carcinoma was previously considered rarer due to the infrequent occurrence of squamous metaplasia of the ductal epithelium [12]. It is impossible to know whether the diagnosis of squamous cell carcinoma of the SMG is in fact malignancy generated within the parenchyma of the gland itself, or possibly a metastatic lymph node adjacent to the SMG, despite the selected code including only primary SMG malignancies. Survival for reported histologic subtypes in our study showed highest rates amongst mucoepidermoid neoplasms, ductal and lobular neoplasms, and adenocarcinomas, and lowest rates amongst epithelial and squamous cell neoplasms. In the current literature, adenoid cystic carcinoma has been shown to have the highest median overall survival at approximately 8–12 years, whereas squamous cell carcinoma exhibited the lowest at 1–4 years [2]. Our research is therefore in line with previously reported rates of malignant transformation regarding various histological subtypes, as well as the aggressiveness of squamous cell cancers relative to other subtypes [12].

Generally, salivary gland tumor incidence increases with age. Ord et al. found malignant submandibular tumors were more common in the 6th decade of life, with the median reported age to be around 55 [12]. This aligns with our results, as almost 30% of those diagnosed were within the 15-year window from 50 to 64 years of age. In addition, almost half of those diagnosed were over 65 with the highest prevalence, in order, among the 7th, 8th, and 6th decades of life. Since a greater percentage of SMG tumors are malignant, they are likely more common in older adults who experience compromises in immune function and increased lifetime exposure to carcinogens. However, our survival analysis found the age group from 50–64 had a lower overall survival compared to patients over 65. Our sample was 54.2% female, deviating slightly from previous study showing a slight predilection for males [2]. Our sample indicated that women had overall survival rates approximately double that of men, similar to Lee et al. who also found female sex as beneficial to survival. In other cancers, it has been documented that women have better overall survival compared to men due to a variety of factors outside of tumor characteristics including lifestyle differences and prevalence of additional comorbidities, however, influences of this sex-based disparity remain unknown [13, 14]. Lastly, our study found race indicating differences in overall survival, with non-Hispanic White patients having poorer overall survival relative to Asian, Black, and Hispanic counterparts. This could also be influenced by numerous factors such as presence of additional co-morbidities or delays in time to treatment.

Survival curves by grade indicate similar outcomes for Grade I/II tumors and Grade III/IV tumors. This lends credence to the idea of tumor grade consolidation into two designations that have been previously used in the literature: low- and high-grade. With regards to low-grade salivary malignancies, one study showed 98.9% 5-year survival and 97.4% at both 10-year and 25-year survival with T1/T2 staging of mucoepidermoid carcinoma [15]. Previously specified histologic subtypes within the low-grade category included acinic cell carcinoma and low-grade polymorphous adenocarcinoma [12]. This aligns with our results indicating adenocarcinomas as well as mucoepidermoid carcinomas as having improved survival. High-grade malignancies include salivary duct carcinoma, carcinosarcoma (malignant mixed tumor), squamous cell carcinoma, oncocytic carcinoma, and small and large cell carcinoma. These cellular profiles are associated with high-risk treatment failure, prompting aggressive multimodality approach [12]. Prior literature has showed hazard of death was significantly increased with higher-grade tumors, particularly among Grade III and IV neoplasms. Haderlein et al. found that high tumor grade is a highly significant predictor not only for a shorter metastasis-free survival, but also for a shorter overall survival [16]. This supports our results, as our analysis showed high-grade epithelial cell and squamous cell neoplasms to have the worst survival.

Spread of disease is another important prognostic factor, with some studies having identified cervical lymph nodes as the best estimator of severity for patients with SMG cancer [4]. Rates of metastasis varied in the literature, ranging from 20 to 50%. [17‐19] Bradley et al. found that rate of distant metastasis in salivary gland cancer ranges from 20 to 40% with a higher rate in high-grade tumors [20]. Other retrospective studies have reported a high rate of distant metastases in high-grade tumors [21, 22]. This may inform lower survival rates in higher-grade tumors, as these patients were more likely to already have distant metastasis at time of diagnosis.

Treatment recommendations for salivary gland malignancies are typically extrapolated from parotid tumor data as follows: surgical resection of malignancy primarily up-front followed by pathology driven adjuvant radiation with the potential to add chemotherapy sensitization. Thus, there remains an opportunity to create informed treatment guidelines for SMG tumors specifically, as the behavior of parotid and submandibular gland tumors can vary, with higher rates of malignancy up to 30% in submandibular neoplasms [11, 17]. Differentiation between benign and malignant lesions is challenging using clinical assessment alone; therefore, histological diagnosis is done either via initial whole tumor resection or via fine-needle aspiration cytology under ultrasound guidance. While needle biopsies can aid in treatment planning and limit additional unnecessary surgery, the higher preponderance of SMG tumors for malignancy should lead to caution and a lower threshold when deciding to undergo whole tumor resection as the primary treatment plan.

Those receiving surgical intervention usually have a less invasive burden of disease with higher likelihood of curative treatment, providing rationale to those undergoing surgical therapy having lower hazard ratios. The mainstay treatment for submandibular gland malignancies is surgical management. Survival outcomes may vary depending on the surgical approach, including the completeness of tumor resection and whether a concurrent neck dissection is performed. Standard management of surgically resected, high-grade carcinoma includes surgical excision of the primary site, as well as neck dissection [12]. Postoperative adjuvant radiation is recommended in high-grade salivary carcinoma, and in cases of close or positive margins [12]. Lower-grade tumor diagnoses generally do not require elective neck dissection [23].

Definitive chemotherapy is mainly restricted to locally recurrent or distant metastases and may not improve survival; it is usually employed as a last-line effort [17, 24‐27]. Those receiving systemic chemotherapy may have had more locally advanced disease or frank distant metastasis making them sub-optimal surgical candidates. Additionally, having larger, more irregular, and poorly differentiated tumors predisposes to less favorable outcomes regardless of treatment modality. When indicated, postoperative chemotherapy usually includes drugs such as cisplatin. Another factor impacting survival is that among older patients, adjuvant chemotherapy/radiation therapy has been shown to carry more toxicity and mortality than radiation alone in comparison to younger patients [28]. The Kaplan–Meier curves in this sample indicated those receiving definitive radiation and/or chemoradiotherapy did not have an improved survival rate, likely due to higher-staged neoplasms with poorer prognoses in these patients. This is somewhat underscored by the results stratified by staging, where the point estimates indicate better survival among earlier stages and worse survival among stages III and IV, albeit both estimates are non-significant. Regarding radiation therapy, recommendations of postoperative radiotherapy include locally advanced tumor stage (pT3/T4), invasion of the bone, perineural spread, peri-nodal spread, close or positive resection margins, high-grade histology, and all adenoid cystic carcinomas except low-grade pT1 with wide margins [28‐31].

The SEER database provides substantial data from a considerable number of patients. However, this study is not without limitations. Limitations include its retrospective nature, which restricts the data that can be leveraged to that which can be extracted from the database. Not all patients were able to be categorized by treatment sequence. Additionally, adenocarcinoma was the most common subtype. While our assumption is that this is including adenoid cystic carcinoma, which is known to be the most common histology subtype in submandibular gland pathology, it is not directly stated [10, 32]. Furthermore, this could include conventional adenocarcinoma in addition to acinic cell carcinoma. It is unknown the extent of reliability and consistency regarding reported attributes of each tumor if these neoplasms were characterized by numerous different pathologists. In the database, regarding each given treatment modality, patients were categorized as “yes” or “no/unknown” which did not allow for sub-stratification between patients who did not receive a given therapy versus those whose status was truly unknown. Additionally, some patients recorded in the database are lost to follow-up. Limited conclusions regarding the efficacy of differing management approaches can be drawn, and further investigation should include a prospective study with varying treatment paradigms in SMG malignancies. Finally, some subgroups contained relatively small sample sizes, leading to unstable or imprecise estimates in the odds of death.

This study augments existing data regarding SMG pathology. In this sample, most favorable survival outcomes were observed amongst mucoepidermoid neoplasms, ductal and lobular neoplasms, and adenocarcinomas, and least favorable survival outcomes were seen amongst epithelial and squamous cell neoplasms. Multi-variate analysis indicated differences in survival by tumor grade, histology, and therapeutic modality, with odds of death shown to be lower with modalities involving surgery. This likely reflects selection of patients with earlier-stage disease, better performance status, and resectable tumors, in whom surgery is often combined with adjuvant radiotherapy with curative intent. In contrast, chemotherapy is more commonly administered to patients with advanced stage, poor performance status, or distant metastases, which may explain the higher observed odds of death in this group. Other types of therapy were generally not associated with odds of death, likely due to the latter patients often being poor surgical candidates with unresectable disease. Due to their uncommon nature, SMG malignancies need further prospective study to determine optimal management and prognosis.