Abstract
Introduction: This study aimed to report the efficacy and safety of adjuvant sequential chemotherapy after definitive treatment of salivary duct carcinoma (SDC) compared with the standard treatment alone (surgery with postoperative radiation therapy). Methods: This was a retrospective study of pathologically confirmed 135 SDC patients (study period 2009 to 2022). After curative surgery and adjuvant radiation therapy, 55 of 135 patients decided to receive additional chemotherapy (OP + RT + chemo group), while 80 opted for surgery and radiation (OP + RT group). Treatment outcomes of overall survival (OS), disease-free survival (DFS), and distant metastasis-free survival (DMFS) were compared using a propensity score matching (PSM) analysis. Results: Adjuvant chemotherapy consisted of three cycles of cisplatin-based regimen, which was well tolerated in most patients with minimal adverse events. Multivariable analyses indicated that the addition of chemotherapy did not improve OS (p = 0.05), DFS (p = 0.386), and DMFS (p = 0.735), although there was a trend toward favoring adjuvant chemotherapy in terms of OS. With PSM analysis, OS (OP + RT + chemo to OP + RT, hazard ratio [HR] = 0.40, 95% confidence interval [95% CI] = 0.12–1.29, p = 0.126), DFS (HR = 0.69, 95% CI = 0.30–1.56, p = 0.367), and DMFS (HR = 0.96, 95% CI = 0.46–1.99, p = 0.903) were not statistically different. Conclusions: Current cisplatin-based adjuvant chemotherapy did not significantly improve treatment outcomes of SDC patients over the surgery and adjuvant radiation. Further development or clinical studies are required to improve the outcomes of SDC, including chemotherapeutic, biomarkers, immune checkpoint inhibitors, or treatment strategies.
Plain Language Summary
Salivary duct carcinoma (SDC) is a rare and aggressive cancer that arises in the salivary glands. While the standard treatment involves surgery followed by radiation therapy, the prognosis remains poor, with many patients experiencing recurrence or distant spread of the disease. This study looked at whether adding chemotherapy to the standard treatment could improve SDC patient outcomes. Researchers analyzed 135 patients treated over 13 years and found that while chemotherapy was safe and well tolerated, it did not significantly improve survival or reduce the risk of recurrence. New approaches, such as targeted therapies or immune treatments, are needed to improve outcomes for patients with this challenging disease.
Introduction
Salivary duct carcinoma (SDC) is a rare malignant tumor originating from the ductal epithelial cells of the salivary glands [1]. The incidence of SDC has gradually increased over the past few decades, estimated at one per 100,000 population, with a higher prevalence in men [2]. It is known as one of the most aggressive tumors within the 21 subtypes of primary salivary gland cancer by the latest World Health Organization classification of Head and Neck Tumor Pathology (5th edition, 2022) [3]. The current standard treatments for SDC are complete surgical resection with adjuvant radiation or chemoradiation for patients with adverse features [4, 5]. Surgery includes resection of the primary sites and appropriate neck lymph node dissection in most cases. Despite the standard treatments, the overall median survival of the SDC patients was reported to be shorter than 4 years after the initial diagnosis, and more than half of the patients died within 5 years [6‒9], with distant metastasis as the most common cause of death or treatment failure [10, 11].
For the current reported adjuvant chemotherapy on SDC, the number of involved cases was very small (less than 10 cases for each study), and the response to single cisplatin-based chemotherapy seemed ineffective [12, 13]. Also, combination regimens, including carboplatin and paclitaxel, carboplatin and docetaxel, or cyclophosphamide with doxorubicin and cisplatin, were reported to show the response rate of 39% to 46% [14‒16]. In cases with human epidermal growth factor receptor-2 (HER-2)-positive SDC, adjuvant chemotherapy adding trastuzumab to paclitaxel, carboplatin combination achieved a 62% response rate [17] and better survival outcomes than without trastuzumab [18], but these benefits have not been proved in HER-2-negative SDC.
In patients with non-small cell lung cancer or triple-negative breast cancer, there have been some evidences indicating that platinum-based chemotherapy following definitive treatment can improve disease-free survival (DFS) or overall survival (OS) [19, 20]. In head and neck squamous cell carcinomas, induction chemotherapy often failed to prove the survival benefit over the concurrent chemoradiation approach [21]. Trial results, however, suggested that adding chemotherapy could reduce the rate of distant metastasis in high-risk patients [22]. In real-world clinical practice in the United States (US) (a study on data from the US National Cancer Database), adjuvant chemotherapy has been increasingly included when managing salivary gland malignancies [23]. This study also demonstrated no survival benefit from adjuvant chemotherapy; however, it did not focus on SDC, but included all types of high-risk salivary gland cancer histology [23]. Considering the high rate of systemic metastasis and the dismal prognosis in SDC patients, we retrospectively reviewed the medical records of SDC patients to determine whether the addition of sequential chemotherapy following completion of the standard treatment (surgery and adjuvant radiation therapy [RT]) improved SDC patients’ outcomes in terms of safety and/or efficacy.
Materials and Methods
Study Cohort
This was a retrospective study that compared the outcomes following standard treatments (surgery and adjuvant RT) with or without chemo. The inclusion criteria for this study were patients who had pathologically confirmed salivary gland origin SDC and underwent curative resection and adjuvant RT at our hospital. A total of 244 patients with SDC who met the inclusion criteria between 2009 and 2022 were identified. The exclusion criteria were recurrent or systemic spread diseases at the time of diagnosis, incomplete standard treatments, previous history of head and neck cancer within 5 years, and/or incomplete medical information due to follow-up loss of shorter than 2 years. With these criteria, 109 patients were excluded and a total of 135 patients were eligible for this study. The tumor-node-metastasis status of each tumor was classified according to the 8th Edition of the American Joint Committee on Cancer (AJCC) Staging Manual [24].
Standard Treatments: Surgery and Adjuvant Radiation Treatment
Preoperative histologic diagnosis using needle aspiration cytology or biopsy was attempted in all patients. Histologic diagnosis of SDC subtype was not clearly made in 76 patients (56.3%), while high-grade carcinoma was diagnosed in 28 (20.7%) and SDC in 31 (22.9%). The primary treatment was adequate surgical resection for the primary lesion. During surgery, frozen biopsy confirmed the primary tumor as high-grade carcinoma in unclear cases preoperatively, leading to concurrent selective neck lymph node dissection. Neck lymph node dissection was performed together electively on the cases without clinical nodal metastasis (lymph node levels 1 to 3) or therapeutically on the cN(+) cases (ipsilateral levels 1 to 5). Based on the surgical pathology information, all patients underwent adjuvant RT, delivering 54–66 Gy by either three-dimensional conformal RT (N = 44) or intensity-modulated RT (N = 91).
Adjuvant Sequential Chemotherapy
At the time of completion of adjuvant RT, information about chemo including the potential benefits and morbidity risk was provided to the patients through the multidisciplinary discussion. Then, decisions for chemo were made by patients. In total, 55 of 135 patients decided to receive additional chemotherapy (OP + RT + chemo group), while 80 patients did not (OP + RT group). The chemotherapy regimens consisted of three (median = 3, range 1–4) cycles of cyclophosphamide, doxorubicin, and cisplatin combination (CAP regimen) in 38 patients, docetaxel and cisplatin combination (DP regimen) in seven, 5-fluorouracil and cisplatin combination (FP regimen) in three, and cisplatin alone in 7 patients, respectively. The chemotherapy was administered within 1 month of RT completion, which was given either weekly or once every 3 weeks. Treatment-related toxicities were monitored based on the Common Terminology Criteria for Adverse Events (CTCAE) version 4.0 [25].
Follow-Up and Outcomes
All patients included in this study had a minimum of 2 years of follow-up, except for those who died within 2 years. The starting point of the follow-up period was the time of completion of adjuvant treatments. The endpoint of follow-up was the time of the last hospital visit or documented event (death, recurrence, or metastasis). Cases without predefined events were censored at the last hospital visit. The mean follow-up period was 46.5 months (range 7 to 227, interquartile range 21 to 73). During the clinical courses, the follow-up protocol was based on the current National Comprehensive Cancer Network (NCCN) guidelines [4].
Statistical Analysis
The initial baseline characteristics were compared between the two groups using the chi-square test or Fisher’s exact test, two-sample t test, or Wilcoxon rank sum test, as appropriate. The clinical variables were age, sex, smoking status, and Eastern Cooperative Oncology Group (ECOG) Performance Status. The reported tumor characteristics were initial pathological tumor stage, resection margin (RM) status, lymphovascular invasion (LVI), perineural invasion (PNI), and extra-nodal extension (ENE) of lymph node metastasis. The events of interest were recurrence (documented local, regional, or systemic metastasis) and death. The OS was defined as the duration of time to death from any causes, DFS was the length of time to recurrence (any type), and distant metastasis-free survival (DMFS) was the length of time to diagnosis of systemic metastasis.
Univariable and multivariable analyses using a Cox proportional hazards model were conducted for OS, DFS, and DMFS in all patients and between the two groups. As the distribution of baseline characteristics was uneven between the two groups, we conducted propensity score matching (PSM) analysis to reduce the bias for potential confounding factors. We matched the propensity scores of the selected variables from the univariable analyses (caliper width = 0.15 for OS, DFS, and DMFS) and built a comparison group of 46 (DFS and OS) or 50 (DMFS) matched pairs of patients between two groups (absolute standardized mean difference 0.0 to 0.57). Then, the outcomes of interest were compared between the two treatment groups, after further adjusting the variables >0.1 in ASMD.
To confirm our results, sensitivity analyses were also conducted by exchanging the variables included in the PSM analysis. All statistical analyses were performed using R software version 4.3 (R Foundation for Statistical Computing, Vienna, Austria). A two-tailed p value <0.05 was considered statistically significant.
Results
Baseline Characteristics of Enrolled SDC Patients
All patients had high-grade SDC, which was confirmed by surgical pathology. Of 135 patients included in the study, 80 underwent surgery and postoperative RT (OP + RT) and 55 did surgery, postoperative RT, and subsequent chemotherapy (OP + RT + chemo) (Table 1). The distributions of age, sex, ECOG Status, and smoking status were comparable between the two groups. Although the pathological status (pT and pN) was similar in the two groups, the proportion of patients with tumor-node-metastasis III–IV stages was higher in OP + RT + chemo group than OP + RT group (81.8% vs. 58.7%, p = 0.041). Regarding the adverse pathological features, RM status, PNI, and ENE were not different between groups; however, LVI was more frequent (marginally) in OP + RT + chemo group (54.6% vs. 37.5%, p = 0.050).
Baseline characteristics and treatment outcomes of enrolled SDC patients
N (%) . | Total patients (N = 135) . | OP + RT (N = 80) . | OP + RT + chemo (N = 55) . | p value . |
---|---|---|---|---|
Age (mean±SD) | 61.0±11.7 | 62.5±12.0 | 58.7±11.0 | 0.061 |
Sex (M:F) | 0.728 | |||
Male | 106 (78.5) | 62 (77.5) | 44 (80.0) | |
Female | 29 (21.5) | 18 (22.5) | 11 (20.0) | |
ECOG status | 0.172 | |||
0 | 40 (29.6) | 16 (20.0) | 24 (43.6) | |
1 | 79 (58.5) | 52 (65.0) | 27 (49.1) | |
2 | 16 (11.9) | 12 (15.0) | 4 (7.3) | |
Smoking status | 0.745 | |||
None + ex-smoker | 71 (52.6) | 43 (53.8) | 28 (51.0) | |
Current smoker | 64 (47.4) | 37 (46.2) | 27 (49.0) | |
TNM stage (AJCC 8th ed.) | ||||
pT status | 0.912 | |||
T1 | 22 (16.3) | 13 (16.3) | 9 (16.4) | |
T2 | 43 (31.9) | 27 (33.7) | 16 (29.1) | |
T3 | 40 (29.6) | 22 (27.5) | 18 (32.7) | |
T4 | 30 (22.2) | 18 (22.5) | 12 (21.8) | |
pN status | 0.447 | |||
N0 | 65 (48.2) | 43 (53.8) | 22 (40.0) | |
N1 | 9 (6.7) | 5 (6.3) | 4 (7.3) | |
N2 | 22 (16.3) | 11 (13.7) | 11 (20.0) | |
N3 | 39 (28.9) | 21 (26.2) | 18 (32.7) | |
Overall stage (TNM) | 0.041 | |||
I | 14 (10.4) | 10 (12.5) | 4 (7.3) | |
II | 29 (21.5) | 23 (28.8) | 6 (10.9) | |
III | 22 (16.3) | 11 (13.7) | 11 (20.0) | |
IV | 70 (51.8) | 36 (45.0) | 34 (61.8) | |
RM status | 0.188 | |||
Negative | 42 (31.1) | 28 (35.0) | 14 (25.5) | |
Close <5 mm | 23 (17.0) | 10 (12.5) | 13 (23.6) | |
Positive | 70 (51.9) | 42 (52.5) | 28 (50.9) | |
LVI present | 60 (44.4) | 30 (37.5) | 30 (54.6) | 0.050 |
PNI present | 50 (37.0) | 28 (35.0) | 22 (40.0) | 0.554 |
ENE present | 40 (29.6) | 22 (27.5) | 18 (32.7) | 0.513 |
Chemotherapy regimen | ||||
CAP | 38 (69.1) | |||
DP | 7 (12.7) | |||
Cisplatin (alone) | 7 (12.7) | |||
FP | 3 (5.5) |
N (%) . | Total patients (N = 135) . | OP + RT (N = 80) . | OP + RT + chemo (N = 55) . | p value . |
---|---|---|---|---|
Age (mean±SD) | 61.0±11.7 | 62.5±12.0 | 58.7±11.0 | 0.061 |
Sex (M:F) | 0.728 | |||
Male | 106 (78.5) | 62 (77.5) | 44 (80.0) | |
Female | 29 (21.5) | 18 (22.5) | 11 (20.0) | |
ECOG status | 0.172 | |||
0 | 40 (29.6) | 16 (20.0) | 24 (43.6) | |
1 | 79 (58.5) | 52 (65.0) | 27 (49.1) | |
2 | 16 (11.9) | 12 (15.0) | 4 (7.3) | |
Smoking status | 0.745 | |||
None + ex-smoker | 71 (52.6) | 43 (53.8) | 28 (51.0) | |
Current smoker | 64 (47.4) | 37 (46.2) | 27 (49.0) | |
TNM stage (AJCC 8th ed.) | ||||
pT status | 0.912 | |||
T1 | 22 (16.3) | 13 (16.3) | 9 (16.4) | |
T2 | 43 (31.9) | 27 (33.7) | 16 (29.1) | |
T3 | 40 (29.6) | 22 (27.5) | 18 (32.7) | |
T4 | 30 (22.2) | 18 (22.5) | 12 (21.8) | |
pN status | 0.447 | |||
N0 | 65 (48.2) | 43 (53.8) | 22 (40.0) | |
N1 | 9 (6.7) | 5 (6.3) | 4 (7.3) | |
N2 | 22 (16.3) | 11 (13.7) | 11 (20.0) | |
N3 | 39 (28.9) | 21 (26.2) | 18 (32.7) | |
Overall stage (TNM) | 0.041 | |||
I | 14 (10.4) | 10 (12.5) | 4 (7.3) | |
II | 29 (21.5) | 23 (28.8) | 6 (10.9) | |
III | 22 (16.3) | 11 (13.7) | 11 (20.0) | |
IV | 70 (51.8) | 36 (45.0) | 34 (61.8) | |
RM status | 0.188 | |||
Negative | 42 (31.1) | 28 (35.0) | 14 (25.5) | |
Close <5 mm | 23 (17.0) | 10 (12.5) | 13 (23.6) | |
Positive | 70 (51.9) | 42 (52.5) | 28 (50.9) | |
LVI present | 60 (44.4) | 30 (37.5) | 30 (54.6) | 0.050 |
PNI present | 50 (37.0) | 28 (35.0) | 22 (40.0) | 0.554 |
ENE present | 40 (29.6) | 22 (27.5) | 18 (32.7) | 0.513 |
Chemotherapy regimen | ||||
CAP | 38 (69.1) | |||
DP | 7 (12.7) | |||
Cisplatin (alone) | 7 (12.7) | |||
FP | 3 (5.5) |
SDC, salivary duct carcinoma; OP, surgery, curative; RT, radiation therapy; chemo, adjuvant chemotherapy; SD, standard deviation; ECOG Status, Eastern Cooperative Oncology Group Performance Status; TNM stage, tumor-node-metastasis prognostic stage; AJCC 8th ed., American Joint Committee on Cancer staging system 8th edition; CAP, cisplatin, doxorubicin, and cyclophosphamide; DP, docetaxel and cisplatin; FP, 5-fluorouracil and cisplatin.
Adverse Events of Chemotherapy
In the 55 patients in OP + RT + chemo group, the most common adverse event was grades 1–2 vomiting (7, 12.7%), followed by grades 1–2 laryngeal inflammation (edema) (4, 7.3%), and grade 3 neutropenia (4, 7.3%), respectively (Table 2). Most of the adverse events following chemotherapy were managed conservatively, including hydration, nutritional support, anti-emesis drugs, or short-term steroid treatment. Serious adverse events did not occur, and 52 of 55 patients (94.5%) could complete the scheduled chemotherapy.
Adverse events of chemo (N = 55)
Adverse events . | Regimen . | Grade . | N (%) . |
---|---|---|---|
Vomiting | ≤2 | 7 (12.7) | |
CAP | 4 (7.3) | ||
DP | 1 (1.8) | ||
Cisplatin | 1 (1.8) | ||
FP | 1 (1.8) | ||
Laryngeal inflammation | ≤2 | 4 (7.3) | |
DP | 2 (3.6) | ||
Cisplatin | 2 (3.6) | ||
Neutropenia | CAP | 3 | 4 (7.3) |
Weight loss | 2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
Cisplatin | 1 (1.8) | ||
Diarrhea | ≤2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
DP | 1 (1.8) | ||
Mucositis | ≤2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
Cisplatin | 1 (1.8) | ||
Dry skin | CAP | 1 | 1 (1.8) |
Neck edema | FP | 1 | 1 (1.8) |
Adverse events . | Regimen . | Grade . | N (%) . |
---|---|---|---|
Vomiting | ≤2 | 7 (12.7) | |
CAP | 4 (7.3) | ||
DP | 1 (1.8) | ||
Cisplatin | 1 (1.8) | ||
FP | 1 (1.8) | ||
Laryngeal inflammation | ≤2 | 4 (7.3) | |
DP | 2 (3.6) | ||
Cisplatin | 2 (3.6) | ||
Neutropenia | CAP | 3 | 4 (7.3) |
Weight loss | 2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
Cisplatin | 1 (1.8) | ||
Diarrhea | ≤2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
DP | 1 (1.8) | ||
Mucositis | ≤2 | 3 (5.5) | |
CAP | 2 (3.6) | ||
Cisplatin | 1 (1.8) | ||
Dry skin | CAP | 1 | 1 (1.8) |
Neck edema | FP | 1 | 1 (1.8) |
Grade of toxicity according to Common Terminology Criteria for Adverse Events (CTCAE) v4.0.
CAP, cisplatin, doxorubicin, and cyclophosphamide; DP, docetaxel and cisplatin; FP, 5-fluorouracil and cisplatin.
Treatment Outcomes of Enrolled SDC Patients
Without adjustment of any variables, overall oncological outcomes between the two groups were similar with a mean follow-up of 46.5 months (Table 3). Any type of local recurrence (local and combined loco-regional recurrences) occurred in 19 (23.8%) in OP + RT group and 12 (21.9%) in OP + RT + chemo group. Regional recurrence (regional or combined loco-regional recurrence) was found in 21 (26.3%) in OP + RT group and 16 (29.1%) in OP + RT + chemo group. For distant metastasis, a total of 36 (26.7%) cases were reported: 19 (23.8%) in OP + RT group and 17 (30.9%) in OP + RT + chemo group without a statistical difference (p = 0.429). Similarly, the death events (documented) were noted in 23.6%–30.0% of the enrolled patients, which were not different between the two groups (p = 0.440).
Treatment outcomes of enrolled SDC patients
Variables . | Total patients (N = 135) . | OP + RT (N = 80) . | OP + RT + chemo (N = 55) . | p value . |
---|---|---|---|---|
Local recurrence, n (%) | 9 (6.7) | 6 (7.5) | 3 (5.5) | 0.738 |
Regional recurrence, n (%) | 15 (11.1) | 8 (10.0) | 7 (12.7) | 0.781 |
Combined loco-regional recurrence, n (%) | 22 (16.3) | 13 (16.3) | 9 (16.4) | 0.999 |
Distant metastasis, n (%) | 36 (26.7) | 19 (23.8) | 17 (30.9) | 0.429 |
Death, any cause, n (%) | 37 (27.4) | 24 (30.0) | 13 (23.6) | 0.440 |
Follow-up duration (mean±SD), months | 46.5±39.1 | 45.4±42.1 | 48.1±34.9 | 0.696 |
Variables . | Total patients (N = 135) . | OP + RT (N = 80) . | OP + RT + chemo (N = 55) . | p value . |
---|---|---|---|---|
Local recurrence, n (%) | 9 (6.7) | 6 (7.5) | 3 (5.5) | 0.738 |
Regional recurrence, n (%) | 15 (11.1) | 8 (10.0) | 7 (12.7) | 0.781 |
Combined loco-regional recurrence, n (%) | 22 (16.3) | 13 (16.3) | 9 (16.4) | 0.999 |
Distant metastasis, n (%) | 36 (26.7) | 19 (23.8) | 17 (30.9) | 0.429 |
Death, any cause, n (%) | 37 (27.4) | 24 (30.0) | 13 (23.6) | 0.440 |
Follow-up duration (mean±SD), months | 46.5±39.1 | 45.4±42.1 | 48.1±34.9 | 0.696 |
SDC, salivary duct carcinoma; OP, surgery, curative; RT, radiation therapy; chemo, adjuvant chemotherapy; SD, standard deviation.
Multivariable Analyses of Risk Factors for OS, DFS, and DMFS in SDC Patients
To confirm the actual effect of chemo on oncological outcomes, we conducted multivariable analyses of identified risk factors for OS, DFS, and DMFS in total SDC patients. Before multivariable analyses, we pre-checked the proportional hazard assumption of the included variables with Shoenfeld’s individual test and time-dependent plots and found that the relative hazard remained constant over time with different covariate levels.
As shown in Table 4, the univariable analysis for OS revealed that poor ECOG Status, current smoker, pN status, LVI, PNI, and ENE were significant risk factors. A subsequent multivariable analysis for OS showed that poor ECOG status, current smoker, pN1–2, and PNI were independent risk factors for OS; however, the addition of chemo to surgery and postoperative RT had a marginal significance (p = 0.050). ENE was excluded in the multivariable model due to the multicollinearity with pN status (variance inflation factor >60). Likewise, the multivariable analysis for DFS identified poor ECOG Status, pN1–2, positive RM, and PNI as independent risk factors. In this model, treatment modality itself (OP + RT vs. OP + RT + chemo) did not affect DFS. As for DMFS, the addition of chemo did not affect DMFS in SDC patients, while current smoker, pN1–2, and PNI were risk factors for DMFS in a multivariable analysis (online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000543281).
Uni- and multivariable analyses of risk factors for OS and DFS in SDC patients
Variables . | OS . | DFS . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
univariable analyses . | multivariable analysis . | univariable analyses . | multivariable analysis . | |||||||||
HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | |
Age (1-year increase) | 1.03 | 1.00–1.06 | 0.056 | 1.02 | 0.99–1.04 | 0.135 | ||||||
Sex (F to M) | 0.43 | 0.15–1.23 | 0.115 | 0.47 | 0.21–1.03 | 0.059 | ||||||
ECOG status (2 to 0–1) | 5.11 | 2.52–10.35 | <0.001 | 6.72 | 2.71–16.65 | <0.001 | 2.94 | 1.54–5.59 | 0.001 | 3.48 | 1.62–7.49 | 0.001 |
Smoking status (yes to no) | 2.73 | 1.32–5.63 | 0.007 | 2.30 | 1.09–4.85 | 0.029 | 2.23 | 1.28–3.89 | 0.004 | 1.71 | 0.94–3.10 | 0.079 |
TNM stage (AJCC 8th ed.) | ||||||||||||
pT status | ||||||||||||
T1 (reference) | ||||||||||||
T2 | 1.23 | 0.43–3.53 | 0.706 | 0.95 | 0.40–2.27 | 0.909 | ||||||
T3 | 1.07 | 0.36–3.22 | 0.899 | 1.15 | 0.49–2.73 | 0.747 | ||||||
T4 | 1.97 | 0.69–5.62 | 0.203 | 2.04 | 0.89–4.70 | 0.093 | ||||||
pN status | ||||||||||||
N0 (reference) | ||||||||||||
N1–2 | 3.86 | 1.46–10.19 | 0.006 | 11.19 | 1.72–72.85 | 0.016 | 3.98 | 1.90–8.34 | <0.001 | 4.45 | 1.32–15.02 | 0.016 |
N3 | 5.78 | 2.41–13.85 | <0.001 | 4.55 | 0.79–26.10 | 0.089 | 4.74 | 2.37–9.46 | <0.001 | 2.17 | 0.68–6.90 | 0.191 |
Overall stage | ||||||||||||
I (reference) | ||||||||||||
II | 0.24 | 0.04–1.34 | 0.104 | 0.10 | 0.01–0.68 | 0.018 | 0.28 | 0.07–1.19 | 0.084 | 0.14 | 0.03–0.67 | 0.013 |
III | 0.33 | 0.06–1.83 | 0.206 | 0.11 | 0.01–1.19 | 0.069 | 0.54 | 0.14–2.02 | 0.360 | 0.24 | 0.05–1.17 | 0.078 |
IV | 1.79 | 0.62–5.16 | 0.279 | 0.29 | 0.04–2.07 | 0.219 | 2.34 | 0.92–5.94 | 0.073 | 0.59 | 0.14–2.47 | 0.466 |
RM status | ||||||||||||
Negative (reference) | ||||||||||||
Close (<5 mm) | 0.54 | 0.25–1.14 | 0.106 | 0.82 | 0.34–2.00 | 0.670 | 0.69 | 0.37–1.31 | 0.260 | 0.83 | 0.40–1.71 | 0.610 |
Positive | 1.57 | 0.68–3.64 | 0.292 | 1.70 | 0.60–4.84 | 0.321 | 2.51 | 1.27–4.95 | 0.008 | 2.79 | 1.18–6.59 | 0.019 |
LVI | 2.82 | 1.42–5.60 | 0.003 | 0.64 | 0.24–1.74 | 0.383 | 2.64 | 1.52–4.58 | <0.001 | 0.61 | 0.28–1.33 | 0.214 |
PNI | 2.83 | 1.46–5.46 | 0.002 | 2.48 | 1.07–5.76 | 0.035 | 2.85 | 1.67–4.86 | <0.001 | 2.24 | 1.14–4.41 | 0.020 |
ENE | 3.14 | 1.62–6.08 | <0.001 | 2.51 | 1.47–4.30 | <0.001 | ||||||
Treatment modality | ||||||||||||
OP + RT (reference) | ||||||||||||
OP + RT + chemo | 0.70 | 0.35–1.39 | 0.309 | 0.45 | 0.20–1.00 | 0.050 | 1.10 | 0.64–1.88 | 0.738 | 0.77 | 0.42–1.40 | 0.386 |
Variables . | OS . | DFS . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
univariable analyses . | multivariable analysis . | univariable analyses . | multivariable analysis . | |||||||||
HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | HR . | 95% CI . | p value . | |
Age (1-year increase) | 1.03 | 1.00–1.06 | 0.056 | 1.02 | 0.99–1.04 | 0.135 | ||||||
Sex (F to M) | 0.43 | 0.15–1.23 | 0.115 | 0.47 | 0.21–1.03 | 0.059 | ||||||
ECOG status (2 to 0–1) | 5.11 | 2.52–10.35 | <0.001 | 6.72 | 2.71–16.65 | <0.001 | 2.94 | 1.54–5.59 | 0.001 | 3.48 | 1.62–7.49 | 0.001 |
Smoking status (yes to no) | 2.73 | 1.32–5.63 | 0.007 | 2.30 | 1.09–4.85 | 0.029 | 2.23 | 1.28–3.89 | 0.004 | 1.71 | 0.94–3.10 | 0.079 |
TNM stage (AJCC 8th ed.) | ||||||||||||
pT status | ||||||||||||
T1 (reference) | ||||||||||||
T2 | 1.23 | 0.43–3.53 | 0.706 | 0.95 | 0.40–2.27 | 0.909 | ||||||
T3 | 1.07 | 0.36–3.22 | 0.899 | 1.15 | 0.49–2.73 | 0.747 | ||||||
T4 | 1.97 | 0.69–5.62 | 0.203 | 2.04 | 0.89–4.70 | 0.093 | ||||||
pN status | ||||||||||||
N0 (reference) | ||||||||||||
N1–2 | 3.86 | 1.46–10.19 | 0.006 | 11.19 | 1.72–72.85 | 0.016 | 3.98 | 1.90–8.34 | <0.001 | 4.45 | 1.32–15.02 | 0.016 |
N3 | 5.78 | 2.41–13.85 | <0.001 | 4.55 | 0.79–26.10 | 0.089 | 4.74 | 2.37–9.46 | <0.001 | 2.17 | 0.68–6.90 | 0.191 |
Overall stage | ||||||||||||
I (reference) | ||||||||||||
II | 0.24 | 0.04–1.34 | 0.104 | 0.10 | 0.01–0.68 | 0.018 | 0.28 | 0.07–1.19 | 0.084 | 0.14 | 0.03–0.67 | 0.013 |
III | 0.33 | 0.06–1.83 | 0.206 | 0.11 | 0.01–1.19 | 0.069 | 0.54 | 0.14–2.02 | 0.360 | 0.24 | 0.05–1.17 | 0.078 |
IV | 1.79 | 0.62–5.16 | 0.279 | 0.29 | 0.04–2.07 | 0.219 | 2.34 | 0.92–5.94 | 0.073 | 0.59 | 0.14–2.47 | 0.466 |
RM status | ||||||||||||
Negative (reference) | ||||||||||||
Close (<5 mm) | 0.54 | 0.25–1.14 | 0.106 | 0.82 | 0.34–2.00 | 0.670 | 0.69 | 0.37–1.31 | 0.260 | 0.83 | 0.40–1.71 | 0.610 |
Positive | 1.57 | 0.68–3.64 | 0.292 | 1.70 | 0.60–4.84 | 0.321 | 2.51 | 1.27–4.95 | 0.008 | 2.79 | 1.18–6.59 | 0.019 |
LVI | 2.82 | 1.42–5.60 | 0.003 | 0.64 | 0.24–1.74 | 0.383 | 2.64 | 1.52–4.58 | <0.001 | 0.61 | 0.28–1.33 | 0.214 |
PNI | 2.83 | 1.46–5.46 | 0.002 | 2.48 | 1.07–5.76 | 0.035 | 2.85 | 1.67–4.86 | <0.001 | 2.24 | 1.14–4.41 | 0.020 |
ENE | 3.14 | 1.62–6.08 | <0.001 | 2.51 | 1.47–4.30 | <0.001 | ||||||
Treatment modality | ||||||||||||
OP + RT (reference) | ||||||||||||
OP + RT + chemo | 0.70 | 0.35–1.39 | 0.309 | 0.45 | 0.20–1.00 | 0.050 | 1.10 | 0.64–1.88 | 0.738 | 0.77 | 0.42–1.40 | 0.386 |
ENE variable was excluded in the multivariable models due to the multicollinearity with pN status variable.
OS, overall survival; DFS, disease-free survival; SDC, salivary duct carcinoma; HR, hazard ratio; 95% CI, 95% confidence interval; ECOG Status, Eastern Cooperative Oncology Group Performance Status; TNM stage, tumor-node-metastasis prognostic stage; AJCC 8th ed., American Joint Committee on Cancer staging system 8th edition; OP, surgery, curative; RT, radiation therapy; chemo, adjuvant chemotherapy.
Comparison of Outcomes with PSM Analysis according to Treatment Groups
Based on the univariable analyses, we conducted PSM analysis for OS and DFS and we chose a well-fitted model with a caliper width of 0.15, leaving 46 cases in each treatment group (Table 5). After PSM, we further adjusted the variables with the absolute value of standardized mean difference of more than 0.10 and then compared OS and DFS of post-PSM patients (Table 6). PSM analysis for DMFS resulted in 50 patients in each treatment group, when caliper width = 0.15 (online suppl. Table S2).
PSM of the two groups for OS and DFS comparison (caliper width = 0.15)
Variables . | OP + RT (N = 46), N (%) . | OP + RT + chemo (N = 46), N (%) . | ASMD . | |
---|---|---|---|---|
pre-matching . | post-matching . | |||
Age (mean±SD) | 62.0±11.5 | 60.6±10.4 | 0.33 | 0.13 |
Sex (M:F) | 37:9 (80.4:19.6) | 37:9 (80.4:19.6) | 0.06 | 0.00 |
ECOG status (0–1:2) | 42:4 (91.3:8.7) | 43:3 (93.5:6.5) | 0.25 | 0.08 |
Smoking status (none + ex-smoker:current smoker) | 22:24 (47.8:52.2) | 22:24 (47.8:52.2) | 0.06 | 0.00 |
TNM stage (AJCC 8th ed.) | ||||
pT status (T1:T2:T3:T4) | 8:10:17:11 (17.4:21.7:37.0:23.9) | 7:11:16:12 (15.2:23.9:34.8:26.1) | 0.13 | 0.09 |
pN status (N0:N1–2:N3) | 19:11:16 (41.3:24.0:34.7) | 20:12:14 (43.5:26.1:30.4) | 0.28 | 0.09 |
Overall stage (I:II:III:IV) | 5:5:8:28 (10.9:10.9:17.4:60.8) | 4:6:9:27 (8.7:13.0:19.6:58.7) | 0.53 | 0.11 |
RM status (negative:close <5 mm:positive) | 15:26:5 (32.6:56.5:10.9) | 13:22:11 (28.3:47.8:23.9) | 0.32 | 0.35 |
LVI (no:yes) | 27:19 (58.7:41.3) | 21:25 (45.6:54.4) | 0.35 | 0.26 |
PNI (no:yes) | 29: 17 (63.0: 37.0) | 27: 19 (58.7: 41.3) | 0.10 | 0.09 |
ENE (no:yes) | 29:17 (63.0:37.0) | 32:14 (69.6:30.4) | 0.11 | 0.14 |
Variables . | OP + RT (N = 46), N (%) . | OP + RT + chemo (N = 46), N (%) . | ASMD . | |
---|---|---|---|---|
pre-matching . | post-matching . | |||
Age (mean±SD) | 62.0±11.5 | 60.6±10.4 | 0.33 | 0.13 |
Sex (M:F) | 37:9 (80.4:19.6) | 37:9 (80.4:19.6) | 0.06 | 0.00 |
ECOG status (0–1:2) | 42:4 (91.3:8.7) | 43:3 (93.5:6.5) | 0.25 | 0.08 |
Smoking status (none + ex-smoker:current smoker) | 22:24 (47.8:52.2) | 22:24 (47.8:52.2) | 0.06 | 0.00 |
TNM stage (AJCC 8th ed.) | ||||
pT status (T1:T2:T3:T4) | 8:10:17:11 (17.4:21.7:37.0:23.9) | 7:11:16:12 (15.2:23.9:34.8:26.1) | 0.13 | 0.09 |
pN status (N0:N1–2:N3) | 19:11:16 (41.3:24.0:34.7) | 20:12:14 (43.5:26.1:30.4) | 0.28 | 0.09 |
Overall stage (I:II:III:IV) | 5:5:8:28 (10.9:10.9:17.4:60.8) | 4:6:9:27 (8.7:13.0:19.6:58.7) | 0.53 | 0.11 |
RM status (negative:close <5 mm:positive) | 15:26:5 (32.6:56.5:10.9) | 13:22:11 (28.3:47.8:23.9) | 0.32 | 0.35 |
LVI (no:yes) | 27:19 (58.7:41.3) | 21:25 (45.6:54.4) | 0.35 | 0.26 |
PNI (no:yes) | 29: 17 (63.0: 37.0) | 27: 19 (58.7: 41.3) | 0.10 | 0.09 |
ENE (no:yes) | 29:17 (63.0:37.0) | 32:14 (69.6:30.4) | 0.11 | 0.14 |
OS, overall survival; DFS, disease-free survival; OP, surgery, curative; RT, radiation therapy; chemo, adjuvant chemotherapy; ASMD, absolute value of standardized mean difference; SD, standard deviation; ECOG Status, Eastern Cooperative Oncology Group Performance Status; TNM stage, tumor-node-metastasis prognostic stage; AJCC 8th ed., American Joint Committee on Cancer staging system 8th edition.
Comparison of treatment outcomes between the two groups (OP + RT vs. OP + RT + chemo) after PSM, adjusted with matching variables of ASMD >0.10
Survival outcomes . | Treatment groups . | HR . | 95% CI . | p value . |
---|---|---|---|---|
OS | OP + RT (reference) | |||
OP + RT + chemo | 0.40 | 0.12–1.29 | 0.126 | |
DFS | OP + RT (reference) | |||
OP + RT + chemo | 0.69 | 0.30–1.56 | 0.367 | |
DMFS | OP + RT (reference) | |||
OP + RT + chemo | 0.96 | 0.46–1.99 | 0.903 |
Survival outcomes . | Treatment groups . | HR . | 95% CI . | p value . |
---|---|---|---|---|
OS | OP + RT (reference) | |||
OP + RT + chemo | 0.40 | 0.12–1.29 | 0.126 | |
DFS | OP + RT (reference) | |||
OP + RT + chemo | 0.69 | 0.30–1.56 | 0.367 | |
DMFS | OP + RT (reference) | |||
OP + RT + chemo | 0.96 | 0.46–1.99 | 0.903 |
OP, surgery, curative; RT, radiation therapy; chemo, adjuvant chemotherapy; ASMD, absolute value of standardized mean difference; HR, hazard ratio; 95% CI, 95% confidence interval; OS, overall survival; DFS, disease-free survival; DMFS, distant metastasis-free survival.
Compared with OP + RT group, the hazard ratio of OP + RT + chemo group was 0.40 (95% confidence interval [95% CI] = 0.12–1.29) for OS, 0.69 (95% CI = 0.30–1.56) for DFS, and 0.96 (95% CI = 0.46–1.99) for DMFS, respectively. All survival outcomes showed no statistical difference between the two groups (Fig. 1). Sensitivity analyses with different caliper widths (0.15 or 0.2), different matching variables, and adjustment (or not) consistently indicated that the addition of chemo did not improve OS and DFS in SDC patients over the standard curative surgery and postoperative RT (online suppl. Table S3).
a–f Survival plots for the total SDC patient and PSM cohorts. When comparing the surgery with radiation (OP + RT) and the additional chemotherapy (OP + RT + chemo) groups, there were no statistical differences in OS, DFS, or DMFS between the total SDC patients or PSM cohorts.
a–f Survival plots for the total SDC patient and PSM cohorts. When comparing the surgery with radiation (OP + RT) and the additional chemotherapy (OP + RT + chemo) groups, there were no statistical differences in OS, DFS, or DMFS between the total SDC patients or PSM cohorts.
Discussion
Among SDC patients who undergo the current standard treatments (surgery with postoperative RT), outcomes are not satisfactory: the reported 5-year OS is only 43%–64% [6, 9, 26], requiring additional efforts to improve the outcomes. Our study confirmed a 5-year OS rate of 73.6% and DFS rate of 53.8%–58.8% in SDC patients with distant metastasis rate of 26.7%, which was higher than the loco-regional recurrence rate of 16.3%. These findings align with the previous studies that identified systemic metastasis as the most common cause of treatment failure [10, 11]. Based on these backgrounds, we tried additional sequential chemotherapy to the standard treatments for SDC patients in the curative setting. Chemo has been frequently prescribed in clinical practice in the USA to manage salivary gland malignancies [23]; however, data from the National Cancer Database indicated that it did not provide a survival benefit for high-risk salivary gland cancers [23]. Our findings focusing on SDC histology were also in line with this report. In this study, we compared the impact of additional chemotherapy following the current standard treatments with multivariable analyses, PSM analysis, and multiple sensitivity analyses, and we reached the conclusion that the additional chemotherapy following surgery and RT did not improve OS, DFS, and DMFS in SDC patients.
Regarding the chemotherapeutic agents, the results from prospective randomized clinical trials have been limited due to the rarity of this disease and there still is lack of guidance on chemotherapy usage in SDC [27, 28]. Rather, previous prospective clinical trials with CAP regimen reported response rate of about 38% in the patients with advanced salivary gland carcinoma, who predominantly had adenocarcinoma [29, 30]. Thereafter, cisplatin-based regimens have been commonly chosen for the recurrent or metastatic salivary gland cancer patients, including SDC, with 14%–39% of objective response rate [14, 31, 32]. Similarly, we tried the CAP, DP, FP, or P chemotherapeutic regimens in our study. On average, three cycles of chemo were administered, and no drug-related deaths or grade 4 toxicities were observed during chemotherapy, allowing 94.5% of the patients to complete the scheduled chemotherapy treatments.
However, the previous reports were all based on the outcomes of recurrent or metastatic disease, rather than in the adjuvant settings [14, 28‒31]. Therefore, it is hard to compare the previous results, in which chemotherapy was applied for the patients with gross residual salivary gland cancers, with ours, in which chemotherapy was added in adjuvant setting. With multiple analyses, one thing for sure in our study was that the addition of chemo had a limited role in the prevention of recurrence and systemic metastasis in SDC patients.
Even with our efforts to adjust the confounders to estimate the treatment outcomes, we had several study limitations to draw a more solid conclusion. First, as this study was a retrospective analysis, the allocation to treatment modalities could have been biased. Although we conducted multivariable analyses and PSM analysis to minimize the bias, hidden confounding factors might have existed, which could have affected the treatment outcomes. Second, the chemotherapeutic drugs prescribed in this study were rather heterogeneous, and due to the small number of patients, we could not explore the actual effects according to each chemotherapeutic regimen, if present. Due to the small number of patients in the adjuvant DP, FP, or cisplatin chemotherapy subgroups, the comparison of efficacy among chemotherapeutic regimens was evaluated by dividing the patients into two groups: the CAP group (N = 38) versus the non-CAP group (N = 17). As a result, we found no significant differences in treatment outcomes between these two groups (online suppl. Table S4). Third, this study only included adjuvant sequential chemotherapy rather than concurrent chemoradiation. Therefore, the boosting effect of RT with concurrent chemotherapy remains unanswered in this study. Finally, as the number of patients was not large enough, there might have been a possibility of limitations in drawing the high statistical power. For example, there was a trend toward favoring chemo in terms of OS in our study, despite no statistically significant difference. Given that the small number of patients were analyzed in this study, the interpretation should be made with caution, and the potential benefit of chemotherapy needs to be reevaluated in a large cohort study. Likewise, the pathological N3 and advanced overall stage seemed to be underestimated because of the small patient number or potential interaction with pathological N3 and advanced overall stage in the multivariable analyses for OS, DFS, and DMFS. To overcome such limitations, further investigation through well-designed, controlled prospective research is desired based on our study.
Nevertheless, our study assessed the outcomes of chemo for specific SDC histology among various high-risk salivary gland cancers, which have not been previously reported in real-world setting. This study suggested the future directions to include new chemotherapy regimens (e.g., anti-HER-2 antibody-drug conjugates [33]) based on biomarker studies, immune checkpoint inhibitors, or different treatment strategies (neoadjuvant vs. adjuvant approaches) to improve the outcomes of the SDC patients.
Conclusion
Cisplatin-based chemo, added to the standard surgery with RT, had no additional benefit in improving the treatment outcomes in SDC patients. Further development or clinical studies are required to improve the outcomes of SDC, including chemotherapeutics, biomarkers, immune checkpoint inhibitors, or treatment strategies.
Statement of Ethics
All procedures performed in this study were in accordance with the ethical standards of the institutional and national research committees and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study protocol was approved before data collection, and the need for informed consent was waived by the Institutional Review Board of Samsung Medical Center (Seoul, Korea; Approval No. SMC IRB 2024-04-093) due to the retrospective nature.
Conflict of Interest Statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding Sources
No authors received any funding.
Author Contributions
Y. Kim: investigation, formal analysis, resources, data curation, and writing original draft. N. Choi: methodology, formal analysis, investigation, visualization, and writing – review and editing. E.H. Kim: resources, data curation, and investigation. M.K. Chung, Y.I. Son, and D. Oh: resources, and writing – review and editing. Y.C. Ahn, H.S. Jeong, and M.J. Ahn: conceptualization, investigation, supervision, visualization, and writing – review and editing. S.H. Lee, H.A. Jung, S. Park, and J. Kim: resources, investigation, and writing – review and editing. All authors gave their final approval and agreed to be accountable for all aspects of the work.
Data Availability Statement
The data supporting the findings of this study are available from the corresponding author (H.-S. Jeong) upon request. The data are not publicly available because they contain information that could compromise the privacy of the research participants.