INTRODUCTION — Salivary glands produce and secrete saliva from a glandular structure that includes the secretory acinus, the ducts, and the myoepithelial and basal cells. Salivary gland tumors can be benign or malignant (table 1) and can arise from either the major (parotid, submandibular, and sublingual) or minor salivary glands located in the oral cavity, paranasal sinuses, larynx, and pharynx (figure 1).
The most common malignant salivary gland tumors include adenoid cystic carcinoma (ACC), mucoepidermoid carcinoma, secretory carcinomas, acinic cell carcinoma, and various adenocarcinoma histologies, including salivary ductal carcinoma.
Salivary gland tumors are rare, and information on management of locoregional and metastatic disease is based on retrospective series and phase II trials, most commonly in ACC. These data and the approach to treatment of locoregionally recurrent and metastatic salivary gland tumors are reviewed here.
The initial treatment of locoregional disease and the epidemiology, risk factors, pathology, and clinical features of these tumors are discussed separately. (See "Salivary gland tumors: Treatment of locoregional disease" and "Salivary gland tumors: Epidemiology, diagnosis, evaluation, and staging".)
LOCOREGIONALLY RECURRENT DISEASE — Treatment options for recurrent salivary gland tumors, as with recurrent head and neck cancer in general, are limited by the previous therapy received and the potential resectability of recurrent disease. (See "Treatment of locally recurrent squamous cell carcinoma of the head and neck".)
Treatment of recurrent pleomorphic adenoma (benign mixed tumor) is discussed separately. (See "Salivary gland tumors: Treatment of locoregional disease", section on 'Pleomorphic adenoma'.)
Local relapse — In general, surgical salvage is the optimal treatment for resectable recurrent salivary gland tumors, although outcomes with surgical salvage are disappointing [1,2].
If at all possible, attempts should be made for complete surgical resection since that provides the best chance for locoregional disease control and, potentially, cure. Depending on the location of the recurrence, surgery can be associated with significant morbidity and may involve sacrifice of the skin, facial nerve, or deeper structures of the face and neck. However, these potentially morbid surgeries should be considered in patients without distant metastatic disease as surgery in these patients can be potentially curative.
Even if distant metastatic disease is present, resection of local recurrence may reduce morbidity compared with alternatives, particularly if the patient has already been treated with radiation therapy (RT). For patients who have not received RT as part of treatment for their original disease presentation and who subsequently develop an isolated local recurrence, we suggest adjuvant RT following salvage surgery, as these patients are at high risk for future relapses. If RT has previously been received, reirradiation should be decided on a case-by-case basis in a multidisciplinary setting.
Irradiation or, in some cases, reirradiation is an option for nonsurgical treatment of locally recurrent salivary gland tumors. If reirradiation is being considered, patient selection is critical due to its toxicity and the likelihood of resistant disease, especially with tumor recurrence within six months after the first course of treatment. (See "Reirradiation for locally recurrent head and neck cancer".)
RT for recurrent disease is generally carried out with conventional external beam RT using photons. Alternative approaches include the following:
●High linear energy transfer RT (eg, protons, neutrons, carbon ions) [3]. (See "Salivary gland tumors: Treatment of locoregional disease", section on 'Definitive treatment'.)
●Intraoperative RT (combined with salvage surgery) and brachytherapy offer additional options for salvage therapy but are usually reserved for patients who previously received external beam RT [4]. (See "Reirradiation for locally recurrent head and neck cancer".)
●RT combined with local hyperthermia or concurrent chemoradiotherapy may play a role in the treatment of some locally advanced or recurrent salivary gland tumors, but experience is limited, and these remain investigational approaches [5-8].
If neither surgery nor RT is feasible, then palliative chemotherapy may be appropriate. (See 'Metastatic disease' below.)
Regional recurrence — Neck dissection, with or without postoperative RT, is performed, if feasible, for surgical salvage of regional lymph node recurrence. If the patient has not yet received RT to the neck, postoperative RT may also be used for high-grade tumors, multiple nodes, close margins, or extracapsular lymph node spread. Regional recurrence with adenoid cystic carcinoma (ACC) is a particular problem since the perineural tumor spread is usually intracranial and surgical resection at the skull base is particularly morbid.
METASTATIC DISEASE — The most common sites of distant metastatic involvement with malignant salivary gland tumors are the lung, liver, and bone.
Natural history — The natural history of metastatic disease is variable, and some patients remain asymptomatic for protracted periods of time. This is especially true for adenoid cystic carcinomas (ACCs; particularly when metastases are limited to the lung), acinic cell carcinomas, and myoepithelial carcinomas [9-16]. Although there is a wide spectrum of biologic behavior [17], median survival following the development of metastatic ACC was approximately three years in older series, which is substantially longer than expected for most other solid tumors [10-12]. This is especially true for patients with ACC identified with asymptomatic disease who are diagnosed using screening computed tomography (CT) [12].
Indications for treatment
Local treatments — Patients may have one single site affecting a critical structure (eg, bronchial obstruction from a lung metastasis, pain or fracture from a bony metastasis) but have otherwise stable systemic disease. For these patients, we offer locoregional therapies such as surgery, radiofrequency ablation, or radiation therapy (RT) to the symptomatic site.
Additionally, long-term survival has been reported with definitive local therapy for those with oligometastatic salivary gland tumors, specifically for those with solitary pulmonary or hepatic metastases, with most data in those with ACC histology [14,18-24]. Most of these patients are treated with metastasectomy. However, radiofrequency ablation and RT (including stereotactic body irradiation and brachytherapy) may be of some benefit in ACC and can be alternatives to surgical intervention in select patients.
As an example, in a series of 83 patients with metastatic head and neck cancer who underwent resection of pulmonary metastases, patients with metastatic ACC had a five-year survival of 84 percent [23]. However, overall survival continued to decline until there were no survivors after 14 years. It is unclear if this reflects the indolent nature of this disease or a true benefit of metastasectomy.
Once locoregional therapy is completed with improvement of symptoms, some patients may be observed, while others may benefit from systemic therapy. Indications are discussed below. (See 'Systemic treatment' below.)
Systemic treatment — Systemic therapy, either with chemotherapy or targeted agents, is offered to patients with an increasing rate of systemic disease progression based on imaging studies, with systemic symptoms, with threatened end-organ dysfunction, and/or with a declined performance status due to progressive disease. Systemic therapy can be administered irrespective of whether local treatments have also been administered, but it may be particularly useful when local treatments are not feasible. (See 'Local treatments' above.)
Surveillance for those without treatment indications — For most patients with stable disease or slow disease progression that is minimally symptomatic, we observe until there is clinical evidence of rapid and/or symptomatic disease progression.
There is no standard approach to observation. We offer clinical surveillance with physical examination, laboratory evaluation (including alkaline phosphatase for those with bone metastases and full liver function testing for those with liver metastases), and imaging using a chest radiograph or CT of the chest every three to six months. For tumors with perineural spread noted on pathology, we offer magnetic resonance imaging (MRI) of the head and neck regions. There are no data to support the use of F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT for surveillance.
Approach to systemic therapy — Treatment choice is based on the rate of disease progression and symptoms, histology and presence of clinically actionable molecular targets, and patient performance status and comorbidities.
Chemotherapy is the most common option for systemic therapy. Whether chemotherapy alters the natural history of most metastatic salivary gland tumor histologies, especially ACC, is unclear [25].
Treatments can also be personalized based on actionable molecular targets [25-27]. Therefore, we assess for targetable mutations in patients with recurrent or metastatic salivary gland tumors. Molecular characterization of rare salivary gland tumors is rapidly advancing, and actionable targets, such as neurotrophic tropomyosin receptor kinase (NTRK), human epidermal growth factor receptor 2 (HER2), androgen receptor, and rearranged during transfection (RET) gene fusions. These molecular alterations are identifiable in specific histologies of salivary gland tumors and some have high response rates to targeted therapy [26]. (See 'Molecular testing' below and 'RET fusion-positive tumors' below.)
The goals of treatment are generally palliative since there is no evidence that systemic therapy prolongs survival in metastatic salivary gland tumors, with few exceptions. Palliative and supportive care should be integrated with therapy, similar to other advanced head and neck cancers. (See "Palliative care: Overview of cough, stridor, and hemoptysis in adults" and "Swallowing disorders and aspiration in palliative care: Definition, pathophysiology, etiology, and consequences", section on 'Head and neck cancer'.)
Clinical trial enrollment is encouraged, where available. However, there are few adequate clinical trials to define the optimal approach to patients with metastatic salivary gland tumors because of the rarity of these tumors.
Adenoid cystic carcinoma
Initial therapy — For patients with advanced or metastatic ACC who have stable or indolent, minimally symptomatic disease, we offer either surveillance or local therapies such as surgery, ablative therapies, or RT. (See 'Surveillance for those without treatment indications' above and 'Local treatments' above.)
In our clinical experience, for patients with advanced or metastatic ACC who lack a targetable mutation and meet the clinical indications for systemic therapy, we offer chemotherapy as initial treatment. Unlike other histologies, however, we avoid paclitaxel in those with ACC, due to lack of efficacy. (See 'Systemic treatment' above and 'Choice of chemotherapy' below.)
Treatment for RET-fusion positive tumors of any histology (including ACC) is discussed separately. (See 'RET fusion-positive tumors' below.)
Subsequent therapy — Upon disease progression, for patients with ACC that lack a targetable mutation, we offer subsequent-line treatment with vascular endothelial growth factor (VEGF) tyrosine kinase inhibitors (TKIs), such as lenvatinib, or evaluation for clinical trials. None of these agents is approved for this indication by the US Food and Drug Administration (FDA), and their use is considered off label. We prefer an initial dose reduction of lenvatinib, starting between 10 and 20 mg daily, with uptitration to the standard dose of 24 mg daily as tolerated. In our experience and in clinical trials, initiating lenvatinib at higher doses increased toxicities [28,29].
In a single-arm, open-label phase II trial, 33 patients with ACC and confirmed disease progression received lenvatinib 24 mg orally daily [28]. Patients received prior therapy for recurrent/metastatic disease, including targeted agents (other than lenvatinib), chemotherapy, immunotherapy, or RT. Approximately one-third of the patients in this study had received prior palliative chemotherapy, with or without RT; two-thirds had received adjuvant chemotherapy, with or without RT; and approximately one-third had received targeted therapy, including six patients with VEGF TKIs. After a median follow-up of 5.2 months, lenvatinib demonstrated a partial response rate of 16 percent, a stable disease rate of 75 percent, and median progression-free survival (PFS) of 17.5 months. Grade ≥3 toxicities included hypertension, oral pain, myocardial infarction, posterior reversible encephalopathy syndrome, and intracranial hemorrhage. Approximately two-thirds of the patients required a dose reduction of lenvatinib to between 10 and 20 mg, and approximately one-half discontinued therapy due to treatment-related toxicities.
Although other VEGF inhibitors, such as sorafenib and axitinib, may be acceptable alternatives to lenvatinib for ACC, we prefer lenvatinib because trials of other agents did not require progression as an eligibility criterion, which raises questions regarding the observed rates of clinical benefit. The use of sorafenib, axitinib, and other investigational agents in ACC is discussed separately. (See 'Novel and Investigational options' below.)
Nonadenoid cystic carcinoma
Molecular testing — For patients with nonadenoid cystic tumors eligible for systemic therapy, the choice of chemotherapy versus targeted agents is influenced by the presence of specific genomic alterations. The initial choice of molecular testing is determined based on histology.
For patients with secretory tumors (eg, mammary analog secretory carcinoma [MASC]), we offer assessment for a neurotrophic tyrosine receptor kinase (NTRK) gene fusion. The diagnostic approach to detecting NTRK gene fusions (including other available assays) is discussed separately (algorithm 1). (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Diagnosis'.)
●In those with tumors that test positive for an NTRK gene fusion, we offer first-line therapy with larotrectinib or entrectinib rather than chemotherapy. (See 'Secretory (NTRK gene fusion positive)' below.)
For patients with nonsecretory tumor histology (eg, mucoepidermoid tumors, salivary ductal carcinomas [SDC], adenocarcinomas not otherwise specified [NOS]), we test for both HER2 and androgen receptor status.
●HER2 expression status is assessed using immunohistochemistry (IHC), with reflex to fluorescent in situ hybridization (FISH) as clinically indicated. If positive (ie, IHC 3+, or IHC 2+ plus FISH HER2/chromosome 17 [CEP17] ratio ≥2), we offer combined therapy with trastuzumab plus taxane. (See 'HER2 overexpression' below.)
This is analogous to HER2 testing in breast cancer. (See "HER2 and predicting response to therapy in breast cancer", section on 'Testing for HER2 expression'.)
●Androgen receptor status is assessed based on IHC. If positive, we offer chemotherapy as first-line therapy and androgen receptor blockade as second-line therapy. (See 'Androgen receptor positive' below and 'Choice of chemotherapy' below.)
Secretory (NTRK gene fusion positive) — For patients with secretory carcinomas of the salivary glands, also known as MASC, whose tumors test positive for a neurotrophic tyrosine receptor kinase (NTRK) gene fusion, we offer first-line therapy with larotrectinib or entrectinib rather than chemotherapy. (See 'Molecular testing' above.)
MASC is histologically and genetically similar to secretory carcinomas of the breast. Activating gene fusions of the NTRK3 gene are found in virtually all cases of MASC but not in other subtypes of salivary gland tumors. While various NTRK gene fusions have been identified, the most commonly detected gene in MASC is ETV6-NTRK3, which occurs from the balanced chromosomal translocation t(12;15) [26,30-32]. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Diagnosis'.)
Larotrectinib and entrectinib are oral selective inhibitors of NTRK that have demonstrated high response rates in MASC of the salivary gland [33-36]. Both agents are approved by the FDA for solid tumors (including MASC) with an NTRK gene fusion, without a known acquired resistance mutation, that are either metastatic or unresectable with no satisfactory alternative treatments, or that have progressed following treatment. We prefer larotrectinib over entrectinib as initial treatment in NTRK-positive MASC. Larotrectinib has been used to treat more patients with MASC, has had longer follow-up with more durable responses, and has a better toxicity profile compared with entrectinib, which has been associated with cardiac toxicity and skeletal fractures [37]. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Other toxicities specific to entrectinib'.)
●Larotrectinib – In a combined analysis of three phase I to II studies, 159 patients with various malignancies receiving previous chemotherapy were treated with larotrectinib [33,36]. Among the 21 patients with MASC of the salivary glands, objective responses were seen in 18 patients (90 percent); the median duration of response was 35 months [36].
●Entrectinib – In a pooled analysis of three single-arm trials (ALKA, STARTRK-1, and STARTRK-2), the objective response rate to entrectinib among 54 patients with relapsed, advanced or metastatic, NTRK-positive solid tumors was 57 percent [35]. Among the seven patients with salivary gland tumors (all MASC histologies), the overall response rate was 86 percent, and duration of response ranged between 3 and 17 months [37].
Further data on the efficacy and toxicity of larotrectinib and entrectinib in other tumor histologies are discussed separately. (See "TRK fusion-positive cancers and TRK inhibitor therapy", section on 'Treatment with TRK inhibitors'.)
Nonsecretory
HER2 overexpression — Up to 30 percent of mucoepidermoid carcinomas and up to 40 percent of SDC overexpress human epidermal growth factor receptor 2 (HER2) or have ERRB2 (the HER2 proto-oncogene) amplification as detected by FISH [38]. In contrast, expression of HER2 is unusual in ACCs [39-44].
In salivary gland tumors with non-ACC histology and HER2 overexpression, we offer taxane-based chemotherapy in combination with trastuzumab as first-line therapy. We typically use carboplatin plus paclitaxel with trastuzumab. However, a single-agent taxane (eg, docetaxel or paclitaxel) with trastuzumab is also an appropriate option, particularly for those with a borderline performance status, lower burden of disease, or desire to avoid the toxicities of combination chemotherapy. We typically do not offer single-agent trastuzumab. Ado-trastuzumab emtansine (T-DM1) may be offered in the subsequent-line setting after progression on first-line HER2-directed therapy [39-51].
Trastuzumab, alone and in combination with chemotherapy, has demonstrated efficacy in patients with HER2 overexpression in both SDC and mucoepidermoid carcinoma, with some patients demonstrating long-term durable responses [43,45-51]. However, the individual contributions of HER2 inhibitors and chemotherapy are unknown, given that most studies have only evaluated trastuzumab-containing chemotherapy regimens in salivary gland tumors.
For example, in an open-label phase II trial of 57 patients with HER2-positive SDC, at a median follow-up of 28 months, trastuzumab plus docetaxel demonstrated an overall response rate of 70 percent, with median PFS and overall survival of approximately 9 and 40 months, respectively [47]. Grade ≥3 toxicities included febrile neutropenia and anemia. Single-agent trastuzumab has demonstrated activity in a small observational study. Among three patients with mucoepidermoid carcinoma, one had a partial response lasting longer than two years, while two patients with previously progressive SDC had disease stabilization for 26 and 40 weeks, respectively [40,46].
T-DM1 may be an appropriate subsequent-line option after progression on trastuzumab-based therapy [45,52]. In a phase II trial of 10 HER2-overexpressing salivary gland tumors of varying histology previously treated with HER2-targeted or endocrine therapy, T-DM1 demonstrated an overall response rate of 90 percent, including five patients (50 percent) with complete clinical responses [45]. After a median follow-up of 12 months, median duration of response and PFS were not reached. Treatment was well tolerated, with grade 1 to 2 infusion reactions, transaminitis, anemia, and rash, and grade ≥3 thrombocytopenia.
Use of other HER2 agents for malignant salivary gland tumors is investigational. (See 'Molecularly targeted therapy' below.)
Androgen receptor positive — Hormone receptor expression is virtually absent in most malignant salivary gland tumors, with the exception of androgen receptor expression in SDC (approximately 80 to 90 percent of tumors) and adenocarcinoma NOS [26,47,53-55].
●Initial therapy – In salivary gland tumors with non-ACC/nonsecretory histology, with androgen receptor expression, and without HER2 overexpression, we offer chemotherapy as first-line therapy. (See 'Choice of chemotherapy' below.)
●Subsequent therapy
•Leuprolide acetate and bicalutamide – Upon subsequent disease progression, we offer androgen deprivation therapy (ADT) with leuprolide acetate and bicalutamide. In patients with metastatic salivary gland tumors expressing androgen receptor, ADT has shown efficacy and minimal toxicity. In a single-arm, open-label phase II trial, 36 patients with salivary gland tumors (a majority of which were SDC) received combined ADT with leuprolide at 3.75 mg subcutaneously every four weeks plus bicalutamide at 80 mg daily. At a median follow-up of 15 months, ADT demonstrated an objective response rate of approximately 42 percent, median PFS of approximately nine months, and median overall survival of approximately 31 months [54,56,57].
A phase II randomized clinical trial (NCT01969578, European Organisation for Research and Treatment of Cancer [EORTC] 1206) evaluating androgen receptor blockade with leuprolide and bicalutamide versus chemotherapy (cisplatin plus doxorubicin, or carboplatin plus paclitaxel) is ongoing. The results will help to optimize the sequencing of ADT in androgen receptor-expressing salivary gland tumors.
•Enzalutamide – In a phase II trial, enzalutamide demonstrated some limited activity among patients with advanced unresectable and metastatic salivary gland tumors that expressed androgen receptor, including those treated with prior ADT [58].
No molecular target identified — Patients with nonadenoid cystic tumors who lack molecular targets are treated with chemotherapy or may be offered clinical trials. (See 'Choice of chemotherapy' below and 'Novel and Investigational options' below.)
RET fusion-positive tumors — For most patients with a metastatic salivary gland tumor of any histology harboring a rearranged during transfection (RET) gene fusion and no other actionable mutations, we suggest initial therapy with selpercatinib rather than chemotherapy. Although data are limited to treatment-refractory disease, selpercatinib offers the opportunity for a prolonged treatment response, is well tolerated, and can be administered orally for patient convenience [59]. In contrast, the optimal chemotherapy regimen is not established for most salivary gland tumors (including ACC). Chemotherapy also has modest activity with limited durable responses and significant risk of toxicity. (See 'Choice of chemotherapy' below.)
Selpercatinib is also an option for patients with RET fusion-positive tumors who have progressed on chemotherapy or other systemic agents and have not previously received this targeted agent.
RET gene fusions have been seen in solid malignancies such as lung, thyroid, and gastrointestinal cancers, but are rarely observed in salivary gland cancers. Selpercatinib, a selective RET kinase inhibitor, was evaluated in an open-label phase I/II basket trial (LIBRETTO-001) of 41 patients with RET fusion-positive tumor agnostic cancers who had progressed on prior systemic therapies [59]. Within the subgroup of four patients with salivary gland tumors, objective responses were seen in two patients (50 percent) and median duration of response was not reached. Grade ≥3 toxicities included hypertension (22 percent) and increased serum transaminases (13 to 16 percent). However, the true clinical benefit of selpercatinib in salivary gland tumors is unclear due to the limited number of patients and further data are necessary.
Based on these data, the FDA granted accelerated approval to selpercatinib for adult patients with locally advanced or metastatic solid tumors with RET gene fusion that have progressed on or following prior systemic treatment or who have no satisfactory alternative treatment options [60]. The efficacy of selpercatinib in other solid tumors is discussed separately. (See "Medullary thyroid cancer: Systemic therapy and immunotherapy" and "Treatment of small bowel neoplasms" and "Systemic therapy for advanced cholangiocarcinoma", section on 'RET fusion-positive tumors' and "Systemic therapy for nonoperable metastatic colorectal cancer: Approach to later lines of systemic therapy", section on 'RET fusion-positive tumors' and "Second-line systemic therapy for advanced exocrine pancreatic cancer", section on 'RET fusion-positive tumors'.)
Choice of chemotherapy — For those in whom systemic therapy is warranted, chemotherapy is typically chosen as initial therapy (with the exception of those with NTRK fusion-positive cancers). (See 'Secretory (NTRK gene fusion positive)' above.)
For most of those eligible for systemic treatment with chemotherapy, we suggest the use of platinum and doxorubicin combination chemotherapy (such as cisplatin, doxorubicin, and cyclophosphamide [CAP]) as first-line treatment, rather than single-agent chemotherapy, in order to maximize treatment response. However, for those with HER2-positive tumors, we typically opt for a platinum and taxane doublet to pair with trastuzumab, as discussed above. (See 'HER2 overexpression' above.)
However, for some patients, single-agent chemotherapy may be appropriate (eg, those with a borderline performance status). If single-agent therapy is used, we offer vinorelbine or mitoxantrone. We do not use paclitaxel in tumors with ACC histology given the lack of efficacy.
Limited data suggest that there are differences in chemotherapy sensitivity among the histologic subtypes of salivary gland tumors. As an example, single-agent paclitaxel appears to have activity against adenocarcinomas and mucoepidermoid carcinomas but not ACCs [25,61]. Whether there are differences with other agents or regimens is difficult to ascertain because of the small numbers of patients included in individual reports (table 2). The data regarding various regimens are summarized as follows:
●Cisplatin, doxorubicin, and cyclophosphamide (CAP) – The most commonly studied regimen is cyclophosphamide (500 mg/m2 intravenously), doxorubicin (50 mg/m2), and cisplatin (50 mg/m2) administered on day 1 and repeated every 28 days [62]. Reported objective response rates are 40 to 50 percent, with duration of response ranging from three to seven months (table 2) [62-68]. Combination chemotherapy regimens generally result in higher response rates but are not clearly superior in survival outcomes, despite additional toxicity. There are no trials that compare CAP versus single-agent therapy or other combination regimens.
●Lack of efficacy of paclitaxel in adenoid cystic carcinoma – A systematic review failed to find sufficient activity for paclitaxel in the treatment of metastatic or locally recurrent ACC [25].
●Single chemotherapy agents
•Vinorelbine, mitoxantrone, cisplatin, doxorubicin (or epirubicin), and methotrexate have modest single-agent activity in treating metastatic disease, with response rates between 10 and 40 percent (table 2) [25,69-73].
•In ACC, single-agent cisplatin has similar efficacy to mitoxantrone, vinorelbine, and epirubicin, but it is associated with greater toxicity and should not be routinely used.
●Other combination regimens – Other regimens evaluated in patients with advanced or recurrent disease include the following (table 2):
•The addition of fluorouracil to CAP had a 50 percent objective response rate in 16 evaluable patients but was associated with significant toxicity, including two treatment-related deaths (one from neutropenic sepsis and the other from doxorubicin-related cardiotoxicity) [74].
•Cisplatin, doxorubicin (or epirubicin), and fluorouracil (PAF) resulted in objective response rates of 35 to 40 percent in patients with metastatic disease [75-77]. Response duration was 6 to 15 months, and toxicity was tolerable.
•In one of the only randomized trials that were conducted in patients with metastatic disease, 36 patients were randomly assigned to vinorelbine (25 mg/m2 on days 1 and 8) plus cisplatin (80 mg/m2 on day 1) or to vinorelbine alone (30 mg/m2 weekly) [73]. Combination therapy was associated with a significantly higher complete response rate (19 versus 0 percent), a significantly higher objective response rate (44 versus 20 percent), and a greater likelihood of survival beyond one year (38 versus 5 percent). Toxicity, in particular nausea and vomiting, was greater in the cisplatin-containing arm.
•Other platinum-based regimens, such as cisplatin plus mitoxantrone [78], cisplatin plus fluorouracil [9], cisplatin plus gemcitabine [79], carboplatin plus paclitaxel [80,81], and cisplatin plus bleomycin with either methotrexate or doxorubicin [9,63,82,83], have demonstrated moderate activity in small trials (table 2).
NOVEL AND INVESTIGATIONAL OPTIONS
Immunotherapy — Immunotherapy with checkpoint inhibitors has shown limited activity in most patients with salivary gland tumors, and data are too preliminary to recommend the use of these agents outside a clinical trial setting. The exception are those with salivary gland tumors expressing high levels of tumor mutational burden (TMB), for which pembrolizumab has regulatory approval based on the results of KEYNOTE-158. Further details are discussed separately. (See "Tissue-agnostic cancer therapy: DNA mismatch repair deficiency, tumor mutational burden, and response to immune checkpoint blockade in solid tumors", section on 'Tumors with high mutational burden'.)
As an example, in the phase I KEYNOTE-028 trial of 26 patients with salivary gland tumors expressing programmed cell death ligand 1 (PD-L1), after a median follow-up of 20 months, pembrolizumab demonstrated an overall response rate of 12 percent, with median progression-free survival (PFS) and overall survival of 4 and 13 months, respectively [84].
There are ongoing trials evaluating checkpoint inhibitor immunotherapy in patients with metastatic salivary gland tumors, such as nivolumab (Nivolumab in Recurrent or Metastatic Salivary Gland Carcinoma of the Head and Neck [NISCAHN]; NCT03132038).
Molecularly targeted therapy — Most targeted agents have not improved objective response rates in patients with salivary gland tumors. Some of these agents have stabilized disease, but this could reflect either antitumor activity or the indolent natural history of some salivary gland tumors, as many trials did not require evidence of disease progression prior to enrollment.
●Tyrosine kinase inhibitors – Various tyrosine kinase inhibitors (TKIs) targeting vascular endothelial growth factor (VEGF) have been evaluated in adenoid cystic carcinoma (ACC) and other salivary gland tumors, including sunitinib, nintedanib, sorafenib, and axitinib [85-89]. Sorafenib and axitinib have demonstrated clinical response rates of 16 and 9 percent, respectively [87-89]. Sunitinib and nintedanib have demonstrated no objective response rates, only disease stabilization [85,86].
●Other targets – Examples of other targets that have been investigated in clinical trials include the following:
•Epidermal growth factor receptor (EGFR) – EGFR inhibitors, such as gefitinib, cetuximab, and lapatinib, have demonstrated minimal activity, despite EGFR expression in ACCs and mucoepidermoid cancers [90-94].
•Farnesyltransferase – In one phase II study, the farnesyltransferase inhibitor tipifarnib demonstrated disease stabilization in approximately half of 13 patients with recurrent or metastatic HRAS-mutant salivary gland tumors [95].
A number of agents have been investigated in ACCs, in particular:
•c-kit TKIs, such as imatinib and dasatinib, have demonstrated minimal activity, despite c-kit expression in a majority of ACCs [25,96-98].
•Everolimus, a mammalian target of rapamycin (mTOR) inhibitor, demonstrated no objective responses in patients with ACC but did reduce tumor burden based on pre- and on-treatment F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET) [99].
•Bortezomib, a proteasome inhibitor of the nuclear factor kappa B (NF-KB) pathway, demonstrated stable disease without any objective responses in patients with ACC [100,101].
•A phase II trial evaluating the NOTCH inhibitor AL101 (NCT03691207) in ACCs is ongoing.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Head and neck cancer".)
SUMMARY AND RECOMMENDATIONS
●Anatomy – Salivary glands produce and secrete saliva from a glandular structure that includes the secretory acinus, the ducts, and the myoepithelial and basal cells. (See 'Introduction' above.)
Salivary gland tumors can be benign or malignant (table 1) and can arise from either the major (parotid, submandibular, and sublingual) or minor salivary glands located in the oral cavity, paranasal sinuses, larynx, and pharynx (figure 1).
●Locoregionally recurrent disease – For patients with locoregionally recurrent disease, surgical salvage is the optimal treatment if the disease is resectable. (See 'Locoregionally recurrent disease' above.)
For patients who undergo salvage surgery and who have not received radiation therapy (RT) as part of treatment for their original disease presentation, we suggest adjuvant RT rather than observation alone (Grade 2C). (See 'Local relapse' above.)
●Treatment indications (metastatic disease) – In patients with metastatic disease, treatment indications are as follows (see 'Indications for treatment' above):
•For patients who have one single site affecting a critical structure but have otherwise stable systemic disease, or for those with oligometastatic disease, we suggest locoregional therapies, with or without systemic therapy, rather than systemic therapy alone (Grade 2C). Patients with solitary metastases, particularly in the lungs or liver and of adenoid cystic carcinoma (ACC) histology, may be most likely to benefit from this approach. (See 'Local treatments' above.)
•Systemic therapy is offered to patients with an increasing rate of systemic disease progression based on imaging, with systemic symptoms, with threatened end-organ dysfunction, and/or with a declined performance status due to progressive disease. (See 'Systemic treatment' above.)
•By contrast, for patients with indolent, asymptomatic, or minimally symptomatic disease, regardless of histologic subtype, we offer surveillance. (See 'Surveillance for those without treatment indications' above.)
●ACC or non-ACC histology lacking a driver mutation
•Initial therapy – For most patients with indications for systemic treatment and with either ACC or non-ACC histology lacking a molecular driver mutation, we suggest a combination chemotherapy regimen as initial treatment (eg, cisplatin, doxorubicin, and cyclophosphamide [CAP]), rather than single-agent therapy, to maximize treatment response (Grade 2C). However, for some patients who may not be able to tolerate chemotherapy combinations, a single agent may be an appropriate alternative. If single-agent therapy is used, we offer vinorelbine or mitoxantrone. For those with ACC histology, we avoid taxanes in the chemotherapy regimen, due to lack of efficacy. (See 'Adenoid cystic carcinoma' above and 'Choice of chemotherapy' above.)
•Subsequent therapy – For patients with ACC tumors lacking a driver mutation who progress on initial therapy and remain candidates for treatment, we offer subsequent therapy with the vascular endothelial growth factor (VEGF) tyrosine kinase inhibitor (TKI) lenvatinib. Sorafenib and axitinib are alternative options that also target VEGF inhibition. (See 'Adenoid cystic carcinoma' above.)
●Non-ACC histology – For patients with non-ACC histology and indications for treatment, we obtain testing for specific molecular targets based on histology to help determine choice of systemic therapy. (See 'Molecular testing' above.)
•Secretory tumors – For non-ACC secretory tumors (ie, mammary analogue secretory carcinoma), we obtain testing for a neurotrophic tyrosine receptor kinase (NTRK) gene fusion (algorithm 1). For those who test positive for an NTRK gene fusion without a known acquired resistance mutation, we suggest targeted therapy with larotrectinib or entrectinib rather than cytotoxic chemotherapy (Grade 2C). (See 'Secretory (NTRK gene fusion positive)' above and "TRK fusion-positive cancers and TRK inhibitor therapy".)
•Nonsecretory tumors – For non-ACC nonsecretory tumors (eg, mucoepidermoid carcinoma, salivary ductal carcinoma, and adenocarcinoma not otherwise specified), we obtain testing for androgen receptor and human epidermal growth factor receptor 2 (HER2) expression.
-HER2 positive – For patients with overexpression of HER2, we suggest the addition of trastuzumab to taxane-based chemotherapy (Grade 2C). For patients progressing on HER2-directed therapy, we offer ado-trastuzumab emtansine (T-DM1) as subsequent therapy. (See 'HER2 overexpression' above.)
-Androgen receptor positive, HER2 negative – For patients with androgen receptor expression and without HER2 overexpression, we suggest chemotherapy as initial treatment rather than androgen deprivation therapy (ADT) (Grade 2C). However, we opt for ADT with leuprolide plus bicalutamide upon disease progression. Enzalutamide also has some limited activity as subsequent therapy. (See 'Androgen receptor positive' above.)
●RET fusion-positive tumors – For patients with rearranged during transfection (RET) fusion-positive salivary gland tumor of any histology and no other actionable mutations, we suggest initial therapy with selpercatinib rather than chemotherapy or other systemic agents (Grade 2C). (See 'RET fusion-positive tumors' above.)