Introduction
Salivary gland carcinomas represent a rare group of malignancies, accounting for 6–8% of all head and neck cancers, and can involve the major or minor salivary glands. According to the World Health Organization (WHO) classification, there are 24 subtypes of salivary gland carcinomas, whose behavior and clinical course vary based on site of origin and histology. The two major categories of salivary gland carcinomas include tumors arising from the intercalated duct, which tend to be lower grade and more indolent, including adenoid cystic carcinoma (ACC), adenocarcinoma, acinic cell carcinoma, myoepithelial carcinoma, and tumors arising from the secretory duct, including salivary duct carcinoma and mucoepidermoid carcinoma (MEC).
Local and Locoregional Disease
The definitive treatment for resectable disease is complete surgical resection. Any adjuvant therapy is reserved for patients with a higher risk of recurrence based on stage, histology, and grade, with radiation therapy (RT) being the primary modality of adjuvant treatments. Evidence for the role of chemoradiotherapy (CRT) has been limited to single institution retrospective studies, and studies have not shown any efficacy of CRT over RT alone. Furthermore, in a review comparing outcomes of patients who received either CRT or RT alone, 2210 patients from the National Cancer Data Base showed a trend toward worsening 2- and 5-year overall survival in patients who received CRT. RTOG-1008 is a prospective ongoing trial evaluating adjuvant RT with or without chemotherapy in high-risk salivary gland tumors.
Local and Locoregional Disease
The definitive treatment for resectable disease is complete surgical resection. Any adjuvant therapy is reserved for patients with a higher risk of recurrence based on stage, histology, and grade, with radiation therapy (RT) being the primary modality of adjuvant treatments. Evidence for the role of chemoradiotherapy (CRT) has been limited to single institution retrospective studies, and studies have not shown any efficacy of CRT over RT alone. Furthermore, in a review comparing outcomes of patients who received either CRT or RT alone, 2210 patients from the National Cancer Data Base showed a trend toward worsening 2- and 5-year overall survival in patients who received CRT. RTOG-1008 is a prospective ongoing trial evaluating adjuvant RT with or without chemotherapy in high-risk salivary gland tumors.
Recurrent and Metastatic Disease
The most common sites of distant metastases include lymph nodes, lung, liver, and bone. Incidence of metastasis is higher, as expected, in high-grade tumors, with high-grade salivary duct carcinomas developing metastatic disease in ~46% of cases and ACC developing metastases in 25–55% of cases. Local therapy options for limited recurrent disease or carefully selected oligometastatic disease include surgical resection or reirradiation if possible. ACC is notable for its prolonged indolent course, with a median time of 33 months between diagnosis of metastatic disease and death. In such cases, treatment is usually limited to symptomatic disease. Potential treatments may be limited by prior treatment modalities, however, and if local treatment is not possible or the patient develops distant metastases, palliative systemic treatment, including chemotherapy and molecularly targeted therapies, can be considered.
Chemotherapy
There is a lack of prospective clinical trials to guide recommendations for palliative systemic therapy given the relatively low incidence of salivary gland cancers. Furthermore, trials prior to the 1990s included a limited histologic classification of salivary gland cancers, hence there is difficulty in evaluating histologic responses of tumors to treatments in relation to newer classifications. Various combination and single drug therapies have been evaluated in the treatment of salivary gland carcinomas with variable activity. The most studied regimen includes a combination of cyclophosphamide, doxorubicin, and cisplatin (CAP). Retrospective data, pertaining predominantly to ACC and adenocarcinomas, show overall response rates ranging from 27% to 60%, with median duration of response being 5–7 months. Platinum-based therapies make up the backbone in the majority of the combination treatments, including combinations of CAP with gemcitabine, vinorelbine, doxorubicin, and carboplatin with paclitaxel, with small patient series showing response rates ranging from 20% to 44%. However, responses have not been associated with improved overall survival. Furthermore, combination therapy has not been shown to have a definitive survival advantage over single agent therapy but is associated with increased toxicity. In one of the few prospective trials for salivary gland cancers, a phase II trial randomized 36 patients to single agent vinorelbine to vinorelbine with CAP. Although there was increased response rates seen in the combination group, there were also increased rates of grades 2–3 toxicity. Meanwhile, there was no significant survival benefit in the combination group.
Histology plays an important role in determining sensitivity to chemotherapy regimens. Regimens including CAP, doxorubicin, and 5-FU appear to have increased activity in adenocarcinoma-like cancers. For example, there were similar rates of response to the CAP regimen with ACC, adenocarcinoma, acinar cell carcinoma, and malignant mixed tumors. In turn, high-grade MEC and squamous cell carcinomas displayed similar sensitivity to treatments, such as CAP containing regimens and methotrexate. Among 16 patients treated with methotrexate, there was approximately a 40% response rate among patients with MEC, and no responses seen in those with ACC or adenocarcinoma. Single agent CAP and paclitaxel have also demonstrated activity against MECs, with a response rate of ~20%. The phase II trial of 25 patients treated with single agent paclitaxel showed activity in MEC and adenocarcinoma, with a 26% overall response rate. However, there were no responses seen in ACC. In turn, vinorelbine, in combination and as a single agent, has shown activity in patients with ACC and adenocarcinoma. Other single agent treatments that have been studied in patients with ACC include mitoxantrone and epirubicin, with response rates ranging from 5% to 10%.
Molecular Targeted Therapy
HER2/neu
HER2/neu has emerged as an important target in breast cancer therapy, and has also emerged as a potential target in the treatment of salivary gland carcinomas. The HER2 oncogene encodes for a transmembrane receptor that is part of the epidermal growth factor receptor (EGFR) family, where activation induces a downstream signal cascade that can lead to increased metastatic potential, tumorigenicity, cell growth, and transformation. The largest series of 135 patients evaluating the expression of HER2/neu found increased frequency in the expression of tumors of an excretory duct origin, including salivary duct carcinomas, MECs, and squamous cell carcinomas, with an overall frequency of 55%. There was an especially high expression in salivary duct carcinomas, with 85% of these tumors evincing evidence of HER2/neu overexpression. In turn, tumors of the intercalated duct origin, such as ACC, adenocarcinoma, and acinic cell carcinoma, had an overall frequency of HER2/neu overexpression at 7%. Smaller series have shown HER2/neu expression ranging from 30% to 38% in MEC and 38–80% in salivary duct carcinomas. Expression of HER2/neu in salivary duct carcinomas is associated with increased rates of recurrence, metastases, and decreased survival rates, independent of tumor grade, size, or nodal status.
Trastuzumab, a humanized monoclonal antibody that targets the extracellular domain of HER2, has shown clinical activity in salivary gland carcinomas that overexpress HER2. The only phase II study evaluating the efficacy of trastuzumab in advanced or metastatic salivary gland carcinomas in 135 patients was closed early due to low expression of HER2 in the accrued patients. However, there were 14 patients that were enrolled and received treatment, with one patient with MEC achieving a partial response (PR) for >2 years, and two patients with salivary duct carcinoma achieving disease stabilization for 26 weeks and 40 weeks, respectively. Case studies of patients with HER2 overexpression treated with trastuzumab-based therapy showed durable responses of stable disease to complete response. In patients with MECs or salivary duct carcinomas, it would be reasonable to determine HER2 expression as a potential treatment target.
C-kit
About 80–100% of ACC express c-kit, a transmembrane tyrosine kinase receptor that regulates cell growth and differentiation. It has been implicated in multiple tumor types, including mast cell leukemia, germ cell tumor, small cell lung cancer, gastrointestinal stromal tumors (GISTs), breast cancer, and angiosarcoma. In more indolent growing tubular and cribriform subtypes, staining for c-kit is limited to the inner epithelial membrane, as opposed to in more aggressive solid subtypes, where staining for c-kit has a more diffuse expression. However, no mutations have been found in association with c-kit expression and they do not appear to be the driving factor of c-kit activation.
Imatinib is a tyrosine kinase inhibitor (TKI) with significant efficacy in treating CML, AML, and specifically c-kit mutation GIST. It has been studied in several phase II studies and case reports of the use of imatinib in c-kit positive ACC, as well as two phase II studies of imatinib in combination with chemotherapy. Early case studies showed mixed responses to the single agent imatinib. In four phase II trials of imatinib alone, there were three documented objective responses out of 50 patients who received single agent imatinib, with responses ranging from 3 to 14 months. When evaluated in conjunction with CAP, four objective responses were seen out of a total of 42 patients. Furthermore, significant grade 3 and 4 hematologic toxicities were seen with the addition of CAP. A more recent phase II study evaluating dasatinib showed one objective response out of 40 patients with c-kit positive ACC. The overall lack of significant efficacy of c-kit inhibitors in ACC may be linked to the lack of c-kit mutations seen in ACC. Therefore, targeting c-kit overexpression does not correlate with clinical benefit.
Epidermal Growth Factor Receptor
Another potential target in ACC that has been studied is the EGFR, which is associated with uncontrolled cell growth in malignancies. Over 90% of squamous cell carcinomas of the head and neck express EGFR and these carcinomas have shown significant response to cetuximab, an anti-EGFR monoclonal antibody. Evaluation of ACC show 35–85% expression of EGFR. However, a phase II study evaluating the use of cetuximab in the treatment of patients with recurrent or metastatic salivary gland carcinomas, 23 of whom had ACC, showed no objective responses. The role of cetuximab with platinum-based therapy was evaluated in another phase II trial, where nine patients with locally advanced ACC were treated with cetuximab with concomitant CAP and radiation, and 12 patients with metastatic ACC were treated with cetuximab with CAP and 5-FU. The overall response rate of both groups was 40%, with a 42% response in the metastatic group. The median progression-free survival in this group was 13 months, with a median overall survival of 24 months. A group of 37 patients, 18 of whom had ACC, were treated with gefitinib in a phase II trial, with no objective responses. The median progression-free survival of patients with ACC was 4.3 months, with a median overall survival of 25.9 months. Lapatinib, a dual inhibitor of EGFR and erbB2, also failed to show any objective response in 19 patients with ACC. Single agent EGFR inhibitors have failed to show any effective response in ACC. While there may be a small benefit of adding cetuximab with combination chemotherapy, further studies would be needed to evaluate for true survival benefit. Furthermore, treatment in the palliative setting would need to be weighed against the toxicity of a triple-drug combination therapy, where 50% of patients developed a grade 3 or 4 toxicity.
Multitargeted Tyrosine Kinase Inhibitors
There have been a multitude of newer generation TKIs with broader targets of action. Vascular endothelial growth factor (VEGF) is another potential target in ACC, with expression correlating with stage, vascular invasion, recurrence and metastases, endothelial cell motility, and an overall poor prognostic factor for overall survival. However, sunitinib, a small-molecule inhibitor of VEGFR1-3, c-kit, PDGFRA/B, and RET, did not show any objective responses in 13 patients with recurrent and/or metastatic ACC, although 62% of patients (8 out of 13) had stable disease >6 months. There have been two phase II trials evaluating sorafenib, with one trial not showing any objective responses. The second trial showed an 11% response rate among ACC, and a 22% response rate in non-ACC malignancies of the salivary gland. Although there was no correlation between activity and expression of EGFR on the tumor cells, all responders had stromal component rich in PDGFR-B, which suggests an antiangiogenic effect to the tumor microenvironment, as opposed to the tumor itself. Newer agents that have been evaluated with mixed responses include axitinib, which showed a partial response in 9% of patients with ACC, as well as dovitinib and nintedanib, which did not show any objective responses. Stable disease ranged from 65% to 75%. Taken in total, TKI therapy has not shown consistent enough results to be recommended as standard of care therapy.
