Pleomorphic adenoma subtype II with extensive chondromyxoid component (hematoxylin-eosin, original magnification ×60).
Pleomorphic adenoma showing squamous differentiation (hematoxylin-eosin, original magnification ×100).
Alves FA, Perez DEC, Almeida OP, Lopes MA, Kowalski LP. Pleomorphic Adenoma of the Submandibular GlandClinicopathological and Immunohistochemical Features of 60 Cases in Brazil. Arch Otolaryngol Head Neck Surg. 2002;128(12):1400-1403. doi:10.1001/archotol.128.12.1400
The submandibular gland is involved in only 5% to 10% of the salivary gland tumors, and pleomorphic adenoma (PA) is the most common tumor affecting it. This study describes the clinicopathological features and immunohistochemical expression of Ki-67 and p53 in 60 cases of submandibular salivary gland PAs. Most of the patients were in the third and fifth decades of life and 37 (62%) of them were women. Tumor sizes varied from 1 to 10 cm and the mean time between symptom onset and treatment was 52 months. Only 1 patient experienced local recurrence, 3 years after treatment. Histologically, most tumors consisted chiefly in a chondromyxoid stroma. Stroma-rich PAs were larger than stroma-poor ones (P<.02). All PAs were found negative for Ki-67 and p53. These results show that PAs of the submandibular gland are histologically similar to PAs of other salivary glands, and that they have a low proliferative rate and a good prognosis.
Salivary gland tumors, which represent 1% to 4% of all human neoplasias, affect the parotid gland in more than 70% of cases, with the submandibular gland (5%-10%), the sublingual gland (1%), and the minor glands (5%-15%) sharing the rest.1 Pleomorphic adenoma (PA), the most common salivary gland tumor, accounts for 60% to 70% of all parotid gland neoplasias, 40% to 60% of all submandibular gland neoplasias, and 40% to 70% of all minor salivary gland neoplasias.1,2 Pleomorphic adenoma is characterized by great histological diversity, and myoepithelial cells are considered responsible for the production of the extracellular matrix.3,4
The expression of genes related to cell proliferation and oncogenesis seems to be associated with the prognosis of some oral tumors.5 The expression of Ki-67, assessed by immunohistochemical methods, has been correlated with the mitotic activity, histological grade, and clinical behavior of tumors, including salivary gland tumors.5,6 In addition, the mutation of the tumor suppressor gene p53 is the most common genetic alteration found in human cancers.5
Many studies have been conducted on the PAs of the parotid and minor salivary glands, but few on the PAs of the submandibular gland. The objective of this study was to describe the clinical, histopathological, and immunohistochemical features of submandibular salivary gland PAs.
A total of 61 cases of benign tumors of the submandibular salivary gland were reviewed. Specimens and clinical data were obtained from the Cancer Hospital AC Camargo, São Paulo, Brazil. The study was carried out with the approval of the Human Research Ethics Committee of the hospital. With the exception of a basal cell adenoma, all cases were diagnosed as PAs. The paraffin-embedded specimens were stained with hematoxylin-eosin and classified according to the 4 subtypes ranked by Seifert et al (1980)7 as follows: I, stroma accounts for 30% to 50% of the tumor; II, stroma accounts for more than 80%; III, stroma accounts for less than 30%; and IV, stroma accounts for less than 30% but is predominantly made up of a monomorphic epithelial cellular component of plasmacytoid or hyaline appearance.
An immunohistochemical study to detect Ki-67 and p53 was performed on 3-µm paraffin-embedded sections mounted on coated glass slides. Antigen was retrieved in a citrate buffer (pH 6) with the help of a microwave oven (2 cycles of 12 minutes each). Endogenous peroxidase was blocked by incubation in a 0.05% solution of hydrogen peroxide for 30 minutes. The slides were further incubated first in a 1:20 dilution of normal horse serum to reduce nonspecific binding, then overnight at room temperature in a 1:200 dilution of primary antibodies against Ki-67 (clone MIB-1) and a 1:200 dilution of primary antibodies against p53 (clone DO-7). Peroxidase activity was visualized with a 0.02% solution of diaminobenzidine hydrochloride containing a 0.03% solution of hydrogen peroxidase. All slides were counterstained with hematoxylin. After each step, the sections were washed in a phosphate-buffered saline solution. Negative control was obtained using nonimmune serum instead of the primary antibody. Samples of squamous cell carcinoma of the floor of the mouth were used as positive controls. The statistical analysis was performed using a χ2 test. A P value below .05 was considered significant.
Most of the patients (64%) were between the third and fifth decades of life, with a mean ± SD age of 36.3 ± 16.3 years, and most (62%) were women. Tumor sizes varied from 1 cm to 10 cm, with a mean ± SD of 4.08 ± 1.9 cm. The main symptom was a swelling in the submandibular region, and the duration of symptoms before treatment varied from 2 to 240 months, with a mean of 54.6 months (Table 1).
Fifty-eight of the 60 patients were treated surgically. In 54 of them, PAs were primary tumors; however, PAs had recurred locally in the other 4 patients who had been initially treated in another hospital. The 2 remaining patients did not receive any treatment; 1 died of pulmonary infection 1 month after its diagnosis and the other refused treatment and did not return to the hospital. The treatment of the 54 patients having a primary tumor consisted in the resection of both the tumor and the submandibular gland. Two of the 4 patients having a recurrent PA were also treated by excision of the submandibular gland and the tumor. In the remaining 2, only the tumor was resected because the gland had been removed previously.
Three years after surgery, local recurrence was observed in only 1 of the 54 patients who had primary PA. This patient underwent another surgical treatment and has been asymptomatic for 8 years. Three of the 4 patients having recurrent PAs are asymptomatic; however, the fourth one developed high-grade mucoepidermoid carcinoma (MEC) in the parapharyngeal space 20 years after his second surgery and died. The MEC, however, was not considered to be associated with the previous PA. All patients were followed up for a period ranging from less than 1 month to 464.4 months (median, 75.7 months).
The chief component of most of the PAs was a stroma rich in chondromyxoid areas, and the tumors were classified as subtypes I and II (Figure 1). To verify the association between tumor size and histological subtype, we correlated values below and above the tumors' median size (3.5 cm) with the subtypes. Most of the PAs classified as subtype II (70%) were greater than 3.5 cm in diameter; 68% of the subtype I tumors were less than 3.5 cm; and 64% of the subtype III tumors were less than 3.5 cm (Table 2). No correlation was observed between symptom duration and tumor size. Stroma-rich tumors were significantly larger than cellular tumors (χ2 test, P<.02). Necrosis in a hyalinized area was observed in 1 case, squamous differentiation (Figure 2) in 9 cases, and capsular infiltration in 7 cases. The 4 recurrent PAs showed multilocular areas, but without features of malignancy. Immunohistochemical analysis results were considered negative for Ki-67 and p53 in all cases, although some cells were positive for Ki-67 in 7 tumors (<1% of the tumor cells) and positive for p53 in 11 tumors.
One case in our series of 61 cases of benign tumors of the submandibular gland was classified as basal cell adenoma, solid type. The patient was a 58-year-old woman whose main symptom was a 36-month swelling in the submandibular region. Surgical treatment consisted of the resection of the tumor and the submandibular gland. The patient has been asymptomatic for about 10 years.
Pleomorphic adenoma is the most common neoplasia of the salivary glands. Although ulceration and pain sometimes hamper speech and swallowing in cases of benign salivary gland tumors,8,9 our study of 60 cases of submandibular gland PAs showed that these PAs have few clinical signs or symptoms besides painless swelling. We reviewed all cases of benign submandibular salivary gland tumors recorded at the Cancer Hospital AC Camargo, and only 1 case was diagnosed not as PA, but basal cell adenoma. In our study population, PA involved more frequently the parotid gland (63% of cases), then the minor salivary glands (18%), the submandibular gland (10%), and the sublingual gland (0.1%).10 The mean age of the patients at the time of diagnosis was 36.3 years. Chidzonga et al11 also found a higher prevalence of parotid gland PAs and lower prevalences of minor and submandibular gland PAs, but observed no case of sublingual gland PA. Of the 2410 tumors of the salivary glands evaluated by Eveson and Cawson,2 72.9% involved the parotid gland, 14% the minor salivary glands, 10.7% the submandibular gland, and 0.3% the sublingual gland; the site of the remaining 2.2% was unknown. Most of the submandibular gland tumors (59.5%) were PAs and the patients' mean age was 44.5 years. Of the 38 cases of benign tumors of the submandibular gland analyzed by Laskawi et al,12 35 were PAs, 2 were lipomas, and 1 was an hemangioma. The mean age of the PA patients, 47 years, was similar to that in the Eveson and Cawson2 series, which was carried out 1 decade before ours. Chau and Radden8 and Lopes et al9 reported that the mean age of patients with benign minor salivary gland tumors was 42 years and 43.5 years, respectively. According to the Armed Forces Institute of Pathology series,10 the mean patient age for PAs of all the salivary glands combined is 41.2 years. It is also well accepted that women are more frequently affected by PA than men.2,8,9,11 In our series, women represented 62% of the cases.
The present study verifies that the resection of both the tumor and the involved gland is the treatment of choice for benign submandibular gland tumors.2,9 All 54 primary tumors in our series were treated in this manner, and there was only 1 case of recurrence. Of the 4 patients treated for recurrence, 3 are asymptomatic. The fourth one developed a high-grade MEC in the parapharyngeal space 20 years after his second treatment for PA. The tumors probably were not related because the second primary tumor was located at a different place, and its histopathological features revealed a high-grade MEC without areas suggestive of PA.
Pleomorphic adenoma characteristically shows a variable amount of myxochondroid stroma produced by myoepithelial cells.3,4 Chau and Radden8 studied 53 cases of intraoral PA, of which 29 were stroma poor (subtypes III and IV). These authors compared the histological subtypes of PA with nonencapsulated lesions and found no significant differences between them. They also reported that stroma-poor PAs were larger than stroma-rich ones, and suggested that cellular tumors may grow at a faster rate. In our series, most of the tumors were stroma rich (subtypes I and II). Cesinaro et al,13 like us, found a high incidence of the PA subtypes I and II (78.5%) and observed that stroma-rich PAs were significantly larger than other PAs, but they did not describe the localization of the tumors. These findings are in agreement with those of Naeim et al,14 who suggested that PAs are highly cellular in the early stages of their development.
Microscopical findings such as necrosis, nuclear atypia, hyalinization, invasion of adjacent tissue, and increased abnormal mitotic activity have been related with an aggressive behavior or malignant transformation of PA.15,16 Auclair and Ellis16 analyzed 65 mixed tumors with atypical histological features to verify factors indicative of malignant transformation and observed that prominent zones of hyalinization were associated with carcinoma development. In our series, 1 tumor showed necrosis in a hyalinized area but it was encapsulated and there were no other features indicating malignancy. Allen et al17 described in 5 benign tumors of the salivary glands areas of necrosis that they considered caused by ischemia. Another PA characteristic that may lead to histopathological confusion is extensive squamous differentiation, mainly observed in small biopsy or fine-needle aspiration samples. Because of such differentiation, an MEC or squamous cell carcinoma may be misdiagnosed.18
Capsule infiltration is not correlated with malignant transformation, but may play a role in the recurrence of PAs.16 In our study, there was only 1 case of recurrence, after 3 years, of a primary tumor resected at our hospital. This tumor was classified as subtype II and presented capsule infiltration, which could be linked to the recurrence. We could not analyze the histological slides of the first tumor of the other 4 patients who presented for treatment of a recurrent tumor. However, in 2 cases, the recurrence was probably associated with an inadequate resection because the involved gland was not excised.
In our cases, expression of Ki-67 and p53 was found negative in all tumors. Lazzaro and Cleveland6 analyzed the expression of Ki-67 and p53 in benign and malignant intraoral salivary gland tumors. They found all benign tumors negative for Ki-67 but, in some cases, detected a low expression of p53. Tsuji et al5 evaluated immunoreactivity for p53 in 11 PAs and found only 1 positive case. These findings suggest that PAs have a low proliferative index and that mutant p53 is not generally expressed.
In conclusion, PA is the most common tumor in submandibular glands. Clinically, as in other glands, PA of the submandibular gland grows slowly. There was recurrence in only 1 case, 3 years after treatment. Microscopically, in most subtype II cases, PAs were rich in chondromyxoid areas, and in all cases they were found negative for p53 and Ki-67. The larger tumors had a rich stroma whereas smaller tumors were cellular (P<.02).
Accepted for publication June 24, 2002.
This work was supported by FAPESP, São Paulo, Brazil.
Corresponding author and reprints: Luiz P. Kowalski, MD, PhD, Hospital do Câncer–Departamento de Cirurgia de Cabeça e Pescoço e Otorrinolaringologia, R. Prof. Antônio Prudente, 211, CEP 01509-900, São Paulo, Brazil.