Eligible patients were 18 years of age or older, with an ECOG performance status of 0–1. Patients had to be diagnosed with histologically confirmed SGCs of the head and neck or other sites and with one of the following histologic subtypes: MEC, adenocarcinoma, ductal carcinoma, or ACC. Eligible patients had stage IV disease or recurrent tumor incurable by surgery or radiotherapy and at least one measurable lesion according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1. Given the indolent nature of ACC, only patients who experienced progressive disease were eligible to participate in this study. Disease progression was defined as one of the following occurring within 6 months of study entry: an at least 20% increase in radiologically or clinically measurable disease, appearance of new lesions, or deterioration in clinical status. These exclusion criteria were applied only for ACC but not for non-ACC. Patients had to be expected to survive for approximately 12 weeks or longer, and prior systemic chemotherapy was not allowed, except adjuvant chemotherapy or chemotherapy completed more than 6 months prior. Major surgery or radiotherapy had to be completed at least 4 weeks before enrollment. Other requirements were adequate organ function defined as absolute neutrophil count (ANC) ≥ 1,500/mm³, hemoglobin ≥ 10 g/dL, platelets ≥ 100,000/mm³, total bilirubin ≤ 1.5×upper normal limit (UNL), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) ≤ 2.5×UNL (if liver metastases present, AST and AST ≤ 5.0×UNL), and creatinine clearance (CCr) ≥ 50 mL/min or serum creatinine ≤ 1.5×UNL. Exclusion criteria included uncontrolled hypertension, unstable or severe angina, history of myocardial infarction within the last 12 months, uncontrolled arrhythmia, or congestive heart failure with New York Heart Association classification III or IV. Patients with uncontrolled systemic disease such as diabetes mellitus, thyroid disease, or active infection were also excluded. Symptomatic central nervous system malignancies were excluded, except metastases completely resected or irradiated by whole brain radiation therapy or stereotactic radiosurgery. Patients with alcohol abuse were excluded.

Eligible patients were treated with docetaxel 35 mg/m² over 1 hour on days 1 and 8 and cisplatin 70 mg/m² over 1 hour on day 1, with the doses repeated every 21 days. Tumor response was evaluated every 6 weeks until disease progression. Response was evaluated using computed tomography (CT) or magnetic resonance imaging according to RECIST 1.1 criteria. Treatment was continued until disease progression, unacceptable toxic effects, or withdrawal of informed consent. Dose reduction and treatment delay owing to adverse events (AEs) were allowed. However, if the patients required second dose reduction or could not receive the study treatment within 42 days of the beginning of the previous cycle, they discontinued the study permanently.

All toxicities were graded according to the Common Terminology Criteria for Adverse Events scale, ver. 4.0. Patients were treated with full or −1 level dose, which was 25% dose reduction (26 mg/m² of docetaxel and 53 mg/m² of cisplatin). For subsequent cycles, ANC ≥ 1,500/mm³ and platelets ≥ 100,000/mm³ were required. However, the wash-out period should not exceed 2 weeks. If myelotoxicity was recovered partially after the 2-week wash-out period (ANC ≥ 1,500/mm³ or platelets ≥ 100,000/mm³), both drugs were given at −1 level dose. If there was febrile neutropenia or ANC < 500/mm³, chemotherapy was postponed until ANC ≥ 1,500/mm³ and platelets ≥ 100,000/mm³ and both drugs were administered at −1 level dose.

In cases of abnormal renal function (CCr < 50 mL/min), docetaxel was administered as planned, but dose reduction in cisplatin occurred after a one-week delay. If CCr was ≥ 50 mL/min in the next week, cisplatin was administered at the previous dose. If CCr was not recovered within 3 weeks, patients were dropped out from the study. Dose reductions in both drugs were made for grade 2 peripheral neuropathy and grade 3 or 4 mucositis including esophagitis. If grade 3 or 4 neuropathy occurred, treatment was discontinued. If there were mild symptoms including localized cutaneous reaction such as mild pruritus, flushing, or rash, the infusion rate was decreased until recovery from symptoms. In cases of grade 2 hypersensitivity, diphenhydramine 50 mg intravenously with or without dexamethasone 10 mg intravenously was administered with monitoring until resolution of symptoms. Docetaxel infusion was resumed after recovery of symptoms. Depending on the physician’s assessment of the patient, docetaxel infusion should be resumed at a slower rate and then increased incrementally to the initial planned rate. Depending on the intensity of the reaction observed, additional oral or intravenous per-medication with an antihistamine should be given for the next cycle of treatment, and the infusion rate should be decreased initially and then increased to the recommended rate. For severe symptoms or anaphylaxis, such as bronchospasm, generalized urticaria, systolic blood pressure under 80 mm Hg, or angioedema, docetaxel infusion was discontinued immediately and diphenhydramine 50 mg intravenously with or without dexamethasone 10 mg intravenously and/or epinephrine were administered as needed while monitoring patients until resolution of symptoms. These patients would be taken off the study.

The primary endpoint was response rate, and the secondary endpoints included overall survival (OS), progression-free survival (PFS), and safety and toxicity profiles. PFS was defined as the time from the date of treatment to the date of documented disease progression, the date of death from any cause, or the last follow-up, whichever came first. OS was defined as the time from the date of treatment to the date of death from any cause and was censored at the date of the last follow-up visit. Response rate for SGCs after docetaxel single therapy was reported previously to be 0% to 12% according systemic review [15]. Assuming a response rate of 0.20, target error of 0.05, and power of 90%, we required 38 patients according to SWOG Cancer Research And Biostatistics (CRAB). Considering a 10% dropout rate, the total recruitment needed to include 42 patients. The total study period was expected to be about 48 months (36 months for accrual of patients and 12 months for additional follow-up). This time period could be shortened or prolonged depending on patient accrual rate. Survival analysis was performed using the Kaplan-Meier method and compared by the log-rank test. p-values less than 0.05 were considered statistically significant. Statistical analysis was performed using IBM SPSS statistics ver. 27 (IBM Corp., Armonk, NY).

Between April 16, 2014, and October 13, 2020, a total of 41 patients was enrolled in this study from Samsung Medical Center. The median age was 58 years (range, 32 to 73 years), and the male to female ratio was 58.5% to 41.5%. Forty patients had ECOG performance status of 1. The most common histological subtype was ACC (63.4%), followed by ductal carcinoma (24.4%). Two patients were treatment-naïve, and the other 39 patients had received treatment for tumors, including surgery (87.8%), radiation (78.0%), concurrent chemoradiotherapy (9.8%), and systemic chemotherapy (12.2%). The most common site of metastasis was the lung (n=31, 75.6%), followed by bone (n=6, 14.6%) and regional lymph nodes (n=6, 14.6%) (Table 1). Four patients with SGCs and one patient with nasal cavity cancer previously had been treated with systemic chemotherapy. Among them, three patients had been given neoadjuvant (n=1) or adjuvant (n=2) chemotherapy, and the other two had been given palliative chemotherapy, which was completed 6 months or longer before enrollment in this study.

Median treatment duration was 3.9 months (interquartile range [IQR], 3.2 to 4.7) and median number of treatment cycles was 5.5 (range, 3 to 8). The objective response rate (ORR) was 46.3% (n=19/41; 95% confidence interval [CI], 38.4 to 68.9), with three CR (7.3%) and 16 PR (39.0%). According to histologic subtype, 23.1% of patients (n=6/26) with ACC achieved PR including 1 CR, whereas 86.6% (n=13/15) of patients with other subtypes showed PR including two CR (Table 2). In patients with ductal carcinoma or adenocarcinoma, ORR was 79% with one CR (7.7%) and nine PR (69.2%). The best response in each patient is depicted in Fig. 1 and the CT imaging of the patient who achieved CR is presented in Fig. 2.

The median follow-up duration was 18.8 months (range, 3.8 to 60.7 months). In all patients, the median PFS was 9.4 months (95% CI, 8.4 to 10.5) (Fig. 3A), the median duration of response was 6.8 months (IQR, 4.0 to 10.2), and the median OS was 28.2 months (95% CI, 22.7 to 33.6) (Fig. 3B). According to histological subtype, the median PFS and OS were 8.9 months (95% CI, 6.6 to 11.1) and 27.6 months (95% CI, 17.7 to 37.4), respectively, in patients with ACC and 10.5 months (95% CI, 7.8 to 13.2) and 29.3 months (95% CI, 20.7 to 38.0) in patients with other subtypes. These were not significantly different between subtypes (p=0.931 and p=0.919, respectively) (Fig. 4A and B). In patients with ductal carcinoma or adenocarcinoma (n=13), the median PFS and OS were 9.1 months (95% CI, 6.7 to 11.4) and 29.3 months (95% CI, 15.4 to 43.3), respectively. There was no significant difference compared to the results in all patients (p=0.521 and p=0.545).

There were no treatment-related deaths in this study. The most common AE of any grade was alopecia (61.0%), followed by anorexia (43.9%) and diarrhea (31.7%). Grade 3 or 4 AEs included neutropenia (4.9%) and fatigue (4.9%) (Table 3). Treatment schedule was delayed in 16 patients (39.0%) and dose modification was required in 14 patients (34.1%) due to toxicity. Neutropenia (19.5%) was the leading cause of treatment delay, and diarrhea (4.9%) and mucositis (4.9%) were the most common causes of dose reduction (Table 4).