Introduction
Minor salivary gland carcinoma of the oral cavity or oropharynx is rarely encountered in the clinical setting; the prevalence is reported as approximately only 1 to 2 per cent of malignant head and neck tumours.Reference Endo, Kida, Furusaka, Yamada, Iida and Sakai1, Reference Katori, Tsukuda, Ishitoya, Kawano, Takahashi and Endo2 This study analysed the prevalence of these lesions and reviewed their treatment, based on histopathological findings in patients treated at our centre.
Subjects and methods
This retrospective study involved 22 patients with minor salivary gland carcinoma of the oral cavity or oropharynx who were hospitalised for treatment at the otolaryngology department of Kurume University Hospital between 1976 and 2005. The group consisted of 10 men and 12 women, aged between 35 and 74 years (mean age: 62 years). Complete cancer eradication was achieved in 20 of the 22 patients. The observation period ranged from one year six months to 15 years (mean: eight years one month). The tumour–node (TN) classification for the study group is shown in Table I.
Table I Patients' tumour–node classification
Data represent number of patients. T = tumour; N = node
The patients' primary tumour site and histopathological tumour type were investigated and the disease prevalence calculated. Histopathological examination results were used to determine resection area and the need for additional therapy. Histopathological findings were reviewed in terms of: presence or absence of cancer cells in the stump; presence or absence of intravascular infiltration; presence or absence of neural infiltration; and tissue grade. A single investigator assessed all histopathological findings. The cause of death was recorded for all patients who died during the follow-up period. Patients' treatment results were compared with those of patients with squamous cell carcinoma (SCC) in the same region. Likewise, the survival rates of minor salivary gland carcinoma patients were compared with those of patients with SCC in the same region, after classification by disease stage, tumour type and treatment method. The Kaplan–Meier method was used to calculate survival rates, which were classified by the cause of death. The log-rank test was used to test for significant differences.
All patients who achieved a complete cure underwent surgical treatment, and the treatment results as well as the cause of death were reviewed for these subjects. Cyclophosphamide, adriamycin and cis-diamminedichoroplatinum were given as chemotherapy. The hard palate was excluded from investigation in this study, because of the difficulty in distinguishing it from the accessory sinus.
Results
Primary sites, histopathology and prevalence
Table II shows the patients' primary tumour sites and histopathological tumour types. Adenoid cystic carcinoma was the most frequently observed tumour type (13 patients), followed by mucoepidermoid carcinoma (seven patients), adenocarcinoma (one patient) and basal cell carcinoma (one patient). The patients in the study group comprised eight of 362 patients with cancer of the oral cavity (i.e. tongue and floor of the oral cavity; 2 per cent) and 14 of 275 patients with cancer of the oropharynx (5 per cent), who were treated during the same period.
Table II Minor salivary gland carcinomas: primary sites, tumour types and prevalence
Pts = patients; ACC = adenoid cystic carcinoma; MEC = mucoepidermoid carcinoma; AC = adenocarcinoma; BCC = basal cell carcinoma; ant = anterior; lat = lateral; sup = superior
Treatment results
In patients undergoing radical treatment for oropharyngeal minor salivary gland carcinoma, both the five- and 10-year survival rates were 90 per cent; this was superior to survival rates for oropharyngeal SCC patients, although the difference was not statistically significant (p = 0.06; Figure 1).
Fig. 1 Survival rates for patients with oropharyngeal minor salivary gland carcinoma (grey line; n = 12) and oropharyngeal squamous cell carcinoma (black line; n = 261). p = 0.06, comparing the two survival rates.
In patients with oral cavity minor salivary gland carcinoma, the five- and 10-year survival rates were 73 and 37 per cent, respectively. These patients' 10-year survival rate was poor in comparison with that of oral cavity SCC patients, but no statistically significant relationship was observed (Figure 2).
Fig. 2 Survival rates for patients with oral cavity minor salivary gland carcinoma (grey line; n = 8) and oral cavity squamous cell carcinoma (black line; n = 354). The difference between the two rates was not significant.
Figure 3 shows patients' survival rates according to their disease stage. Survival rates for patients with stage IV tumours were poor in comparison with those for patients with stage I, II or III tumours, and a statistically significant relationship was observed (p = 0.04).
Fig. 3 Patients' survival rates according to disease stage: stage I (black line), stage II (light grey line); stage III (dark grey line) or stage IV (stippled line). p = 0.04, comparing stage IV patient survival with the survival of patients with stages I, II or III.
Figure 4 shows patients' survival rates according to their tumour type. The five-year survival rate for patients with adenoid cystic carcinoma was favourable, at 92 per cent, but survival tended to deteriorate after year five. Because of the limited number of patients, no statistically significant relationship was observed between survival rate and different tumour types.
Fig. 4 Patients' survival rates according to tumour type: basal cell carcinoma (grey line; n = 1); adenoid cystic carcinoma (black line; n = 13); mucoepidermoid carcinoma (hatched line; n = 5); and adenocarcinoma (dashed line; n = 1). The differences between these survival rates were not significant.
Figure 5 shows patients' survival rates according to their treatment method. Survival following surgery plus radiotherapy showed a favourable trend, but a statistically significant relationship was not observed.
Fig. 5 Patients' survival rates according to treatment method: surgery + radiotherapy (stippled line; n = 12); surgery (black line; n = 6); or surgery + chemotherapy (grey line; n = 2). The differences between these survival rates were not significant.
Cause of death
Table III shows the relationship between tumour type, tumour stage and cause of death. With respect to cause of death, three patients died of the primary tumour, two died of distant metastasis and one died of other disease. In the 13 patients with adenoid cystic carcinoma, deaths were due to primary tumour (one patient) and distant metastasis (two patients). The primary tumour was the cause of death in one patient with mucoepidermoid carcinoma and one with adenocarcinoma.
Table III Cause of death, by tumour type and stage
Pts = patients; nodal mets = lymph node metastases; ACC = adenoid cystic carcinoma; MEC = mucoepidermoid carcinoma; AC = adenocarcinoma; BCC = basal cell carcinoma
Clinical course by tumour type and histopathology
Table IV shows the clinical course and histopathological findings of patients with mucoepidermoid carcinoma. All surgical margins were negative. Vascular infiltration was observed in three patients and perineural infiltration in two. Case four was a patient receiving secondary therapy after undergoing radiotherapy at another hospital. This patient's tumour grade was high, and both intravascular and perineural infiltration were observed. Chemotherapy was performed, but the patient died due to progression of the primary tumour.
Table IV Patients with mucoepidermoid carcinoma: clinical course and histopathological findings
Pt = patient; no = number; yr = years; T = tumour stage; N = node stage; RT = radiotherapy; CT = chemotherapy; mth = months; intravasc = intravascular; infltrn = infiltration; perineur = perineural; F = female; M = male; lat = lateral; STG = subtotal glossectomy; MND = modified neck dissection; PG = partial glossectomy; SND = selective neck dissection; LWR = lateral wall resection; TG = total glossectomy; NED = no evidence of disease; DOP = dead of primary tumour; DOC = dead of other cause; −= negative or absent; += positive or present; intermed = intermediate
Table V shows the clinical course and histopathological findings of patients with adenoid cystic carcinoma. The surgical margins were positive in two patients and negative in 11. Intravascular infiltration was observed in four patients and perineural infiltration in six. According to the classification of Szanto et al.,Reference Szanto, Luna, Tortoledo and White3 the tumour grade was I in two patients, II in five and III in six.
Table V Patients with adenoid cystic carcinoma: clinical course and histopathological findings
Pt = patient; no = number; yr = years; T = tumour stage; N = node stage; RT = radiotherapy; mth = months; intravasc = intravascular; infltrn = infiltration; perineur = perineural; F = female; M = male; lat = lateral; sup = superior; TG = total glossectomy; SND = selective neck dissection; TR = tumour resection; MND = modified neck dissection; LWR = lateral wall resection; PG = partial glossectomy; DOC = dead of other cause; DODM = dead of distant metastasis; NED = no evidence of disease; AWD = alive with disease; −= negative or absent; += positive or present
Discussion
Minor salivary gland carcinomas of the oral cavity and oropharynx are rare tumours, and treatment decisions are difficult when such tumours are encountered clinically. Surgical treatment is recommended by many authors, but few studies have described the necessity for post-operative treatment, or the indications for such treatment. The current study reviewed treatment approaches in such patients, based on clinical outcomes and histopathological findings.
Crocker et al. Reference Crocker, Cavalaris and Finch4 reported that 10 of 38 cases of minor salivary gland tumour of the oral cavity were malignant, with mucoepidermoid carcinoma being the most prevalent type. Goldblatt and EllisReference Goldblatt and Ellis5 reported that 50 of 55 patients with minor salivary gland tumours of the tongue had malignant disease, and that tumour progression, age and tumour grade were prognostic factors. Endo et al. Reference Endo, Kida, Furusaka, Yamada, Iida and Sakai1 reported the prevalence of different minor salivary gland tumour types; adenoid cystic carcinomas were commonest, followed by mucoepidermoid carcinomas. According to the Japanese clinical statistics 2001 registry,Reference Saikawa, Nigauri and Nishikawa6 minor salivary gland carcinoma accounted for four of 162 cases of oropharyngeal cancer (2 per cent), and 15 of 478 cases of cancer of the oral cavity (3 per cent) (based on data from 29 sites). In the present study, subjects with minor salivary gland carcinoma comprised 14 of 261 patients with oropharyngeal cancer (5 per cent) and eight of 362 patients with oral cavity cancer (2 per cent) treated during the study period. Thus, the total number of patients in the present study was comparable to that reported by the Japanese clinical statistics 2001 registry. With respect to histopathological tumour type, adenoid cystic carcinoma was commonest, followed by mucoepidermoid carcinoma, which is consistent with the results of Endo et al. Reference Endo, Kida, Furusaka, Yamada, Iida and Sakai1 The prevalence of each tumour type should be examined using data from many more studies, given the limited number of patients included in any one study.
Many authors consider tumour stage to be a prognostic factor for the outcome of minor salivary gland carcinoma treatment.Reference Endo, Kida, Furusaka, Yamada, Iida and Sakai1, Reference Goldblatt and Ellis5, Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7 In the present study, the survival rate of patients with stage IV cancer was poor, and a statistically significant difference was observed compared with the survival rates of patients with stage I, II or III tumours. Few previous studies have compared patient survival results for different histopathological tumour types; however, some survival results have been reported for adenoid cystic carcinoma and mucoepidermoid carcinoma (for which patient numbers are greater).Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7–Reference Heidelberger and Batsakis12
Andersen et al. Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7 assessed minor salivary gland carcinoma treatment results by histopathological tumour type, and found significantly better survival rates in patients with adenoid cystic carcinoma, compared with other tumour types. Sur et al. Reference Sur, Donde, Levin, Pacella, Kotzen and Cooper9 reported both the local control rate and the disease-free survival rate to be better in patients undergoing total resection, compared with patients with surgical margin positive histologically. The local control rate is reported to be better following surgery plus radiotherapy, compared with radiotherapy alone. Many authors recommend surgery plus radiotherapy as the preferred treatment method.Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7, Reference Sur, Donde, Levin, Pacella, Kotzen and Cooper9, Reference Goepfert, Giraldo, Byers and Luna10
The current study observed no statistically significant differences between patient survival rates for different histopathological tumour types, because of the limited number of patients with each tumour type; however, patients with adenoid cystic carcinoma tended to have more favourable results. Unfortunately, the survival rate of patients with adenoid cystic carcinoma began to fall five years after treatment. Our results were consistent with those of previous reports. It is again acknowledged that the disease course should be observed over a long period in patients with adenoid cystic carcinoma. Treatment with surgery plus radiotherapy was observed to have a favourable effect on adenoid cystic carcinoma patient survival, compared with other treatments, but the difference was not statistically significant. This may be attributable to the limited number of patients receiving each treatment method, and to the fact that different tumour types were examined together. Results for patients with mucoepidermoid carcinoma and adenoid cystic carcinoma (who were comparatively numerous) were therefore comparatively more influential (including their histopathological findings).
Many reports recommend surgical resection as the basic treatment method for mucoepidermoid carcinoma;Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7, Reference Olsen, Devine and Weiland11, Reference Heidelberger and Batsakis12 furthermore, Olsen et al. Reference Olsen, Devine and Weiland11 stated that cervical lymph node dissection was needed in patients with large tumours. In addition, patient prognosis has been reported to be poor for high grade mucoepidermoid tumours.Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7, Reference Heidelberger and Batsakis12 In the current study, mucoepidermoid tumour in local and cervical areas was controlled by surgical resection (except in one patient who underwent secondary therapy); however, the number of patients thus treated was limited (five patients). Such favourable control rates may be due to the tumour grade being either low or intermediate in our mucoepidermoid carcinoma patients, and to the fact that the surgical margins were negative in all patients. It is not possible to evaluate the need for post-surgical treatment or the observed histopathological findings (such as the presence or absence of intravascular and perineural infiltration) in patients with mucoepidermoid carcinoma, based on the results of this study, because of the limited number of patients. However, surgical resection is considered to be the first choice for mucoepidermoid carcinoma treatment, since the efficacy of radiotherapy has not yet been confirmed, as discussed in other reports.Reference Andersen, Therkildsen, Ockelmann, Bentzen, Schiodt and Hansen7–Reference Heidelberger and Batsakis12
Many researchers have reported that treatment outcomes for adenoid cystic carcinoma were better after surgery plus radiotherapy compared with surgery or radiotherapy alone.Reference Namazie, Alavi and Abemayor8, Reference Sur, Donde, Levin, Pacella, Kotzen and Cooper9, Reference Kessler, Mickel and Calcaterra13 In our centre, post-operative radiotherapy is actively performed, based on our previous study results.Reference Umeno, Miyajima, Mori and Nakashima14
Various reports have been published concerning histopathological findings for adenoid cystic carcinoma.Reference Szanto, Luna, Tortoledo and White3, Reference Umeno, Miyajima, Mori and Nakashima14–Reference Hisa, Yasuda, Tadaki, Nishiyama, Hukushima and Murakami16 Many of these authors consider histopathological tumour grade to be a prognostic factor; the presence of grade III disease, which contains many solid patterns, is regarded as a poor prognostic sign. In the current study, the surgical margins were positive in two patients undergoing post-operative radiotherapy. Neither of these two patients experienced a three- or five-year relapse, following careful observation of the disease course. One of the three adenoid cystic carcinoma patients who did not receive post-operative radiotherapy died due to progression of the primary tumour. This suggests the necessity of post-operative radiotherapy for patients with adenoid cystic carcinoma. With respect to the tumour grade and detailed histopathological findings, three of the six patients with grade III adenoid cystic carcinoma died of their primary tumour. Thus, all the adenoid cystic carcinoma patients who died of their primary tumour had grade III disease; this would appear to indicate that after-treatment is required for grade III adenoid cystic carcinoma patients. Three of the six adenoid cystic carcinoma patients with perineural infiltration died of their primary tumour, while three of the four adenoid cystic carcinoma patients with intravascular infiltration died of distant metastasis. Chemotherapy should therefore be considered for adenoid cystic carcinoma patients with the above findings.
TsukudaReference Tsukuda17 investigated the choice of chemotherapy for adenocarcinoma, and reported the clinical usefulness of cis-diamminedichoroplatinum for adenocarcinoma. However, the efficacy of this anticancer agent could not be confirmed because of the limited number of patients in this study.
The types and indications of chemotherapy in patients with minor salivary gland carcinoma should be further investigated in larger patient series.
