INTRODUCTION

Pleomorphic adenoma or mixed benign tumor is the neoplasm of salivary glands that has the highest incidence in major and minor salivary glands (Chart 1) 1-3. Only occasionally it presents other cervical-facial areas involved, such as the external auditory canal, nose and larynx 4. The deep and diffuse location of the glands in the connective tissue of various sites of the head and neck determines the difficulties in early clinical diagnosis of these neoplasms.

Imaging techniques have shown a key role in defining origin, location and limits of the lesion, especially the CT scan images. Conventional X-rays are limited and useful to screen adjacent lesions to mineralized tissues. Ultrasound (US), despite being non-invasive and easy to perform 5, is useful in differentiating intra and extraglandular lesions 6, but it is also limited in deep spaces of the head and neck region, especially because of bone overlapping, in addition to being an instructor-dependent technique. Computed tomography scan (CT scan) despite allowing visualization of soft tissues at sections do not offer definition of the details, hindering the differential diagnosis with lesions located deep in the region of the head and neck 7, 8. CT scan requires, in most cases, high doses of ionizing radiation and injection of contrast. Magnetic resonance imaging (MRI) is a tomographic technique that provides high definition in soft tissues, allowing the differentiation between intra and extraglandular lesions and better definition between the salivary glands, the lesions and the surrounding tissues. MRI does not use ionizing radiation, being one common diagnostic method indicated for such purpose 9-12. The purpose of the present study was to use the literature review and discuss the importance of different imaging methods, advantages and disadvantages, focusing on the use of MRI for the diagnosis and treatment plan of pleomorphic adenoma of salivary glands, illustrated by two clinical case reports.

LITERATURE REVIEW

The clinical aspects of pleomorphic adenoma normally include ovoid solitary lesion well delimited margins. The damage is mobile, except when in the palate, presenting slow growth and being asymptomatic; the dimensions may vary from few millimeters to many centimeters, more commonly affecting people aged on the 4th and 5th decades of life, with mild predominance in female subjects 7, 13-17.
Malignant neoplasms in minor salivary glands present lower incidence than 2% in all cases of tumors. However, untreated neoplasms in these glands can suffer malignant transformation, with incidence of 5% to 25% of the cases 3, 7, 18, 19.

A histopathological characteristic of the pleomorphic adenoma is its heterogeneity, with cell proliferation of variable number, presenting myoepithelial characteristics, ductiform structures and stroma with chondroid tissue, mixoid, hyaline, fatty and/or bony tissue, with fibrous connective capsule of variable thickness and integrity. Incomplete encapsulation is more common in neoplasm of minor salivary glands, especially along the superficial portion of those located in the palate, below the epithelial surface 7, 8, 14, 15, 17.

Occasionally, the neoplasm can present surrounding extracapsular lesions which, in sections, show pediculated areas of growth external to the main neoplasm. The large sized dimensions, in addition to pediculated areas, can also be present with necrosis, hemorrhage and calcifications or ossifications 7, 8, 14, 15, 17. The presence of areas of necrosis, atypical mitosis, invasiveness and external hyalinization suggest the possibility of malignancy 14, 18. Calcifications inside the mass suggest benign neoplasm or, less commonly, schawannoma or mucoepidermoid carcinoma 7.

Minor salivary gland damages area histologically similar to the ones that affect the major salivary glands, despite the tendency of many of those glands of being predominantly cellular, with minimum of stroma and cartilaginous component 14, 18. Histopathological identification of these tumors in other regions of the body than salivary glands is not always exact, being that immunohistochemistry can significantly contribute to the formulation of the final diagnosis and conduction of the appropriate follow-up4.

It has been described that cells of ductal and solid areas of the pleomorphic adenoma present greater activity than the chondroid areas, being related to possible malignant transformation 20.

Information provided by conventional X-ray in the diagnosis of AP is limited. The methods are used as screening exam for lesions that are adjacent to the mineralized tissues, such as in the cases of hard palate region. In these areas, a radio transparent image is well defined, provably owing to bone reabsorption by compression, generated by the expansion of the neoplasm, which has been occasionally described and can suggest malignant lesion 21. When minor salivary glands are involved, CT images and/or MRI should be the exams of choice 7.

US is a non-invasive, easy technique of imaging for diagnosis5, useful in differentiating between intra and extraglandular lesions 6. Shimizu et al., in a study with US for the diagnosis of parotid tumors, described the sonomorphologic characteristics of AP emphasizing the lobular form as an important characteristic for differential diagnosis of PA 22. Other authors described limitations in the use of US for the diagnosis of salivary gland lesions and suggested that the investigation be made associated with other methods of imaging, especially in lesions larger than 3cm, located in the palate 7, 23. The limitations include: little availability of transducers that make intraoral exam; no visualization of the whole depth of the parotid gland (medial lobe), owing to mandible interposition 5, 6.

In CT scan images, the small lesions have been described as ovoid and homogenous, whereas larger lesions normally have a heterogeneous aspect, with areas of little attenuation that correspond to cystic and necrosis areas. Areas of increased attenuation frequently represent recent hemorrhage and can be clinically associated with sudden increase in size and localized pain. The lobulated contour, especially in larger neoplasms, is again highlighted as an important signal for differential diagnosis. CT scan is the best technique to identify the dystrophic calcifications when present in the lesion 7, 8.

In case of malignancy of the palate region, in which there is invasion and destruction of the bone cortical area, the CT scan is the exam of choice. In most cases, the characteristics of the images reveals large tumors of heterogeneous images, extending beyond the midline, invading the maxillary sinus with bone cortical layer destruction 7, 24, 25.

MRI can present homogeneous or heterogeneous signal and the understanding of the meaning of this signal is essential to have basic notions of the physical principals of imaging capturing, even though it is not the objective of our study. The MRI corresponds to hydrogen proton mapping in different organic tissues, obtained through measures of processed energy by computer and translated in the form of digital images. In this processing, there are two main involved mechanisms - T1 and T2 - resultant from the interaction of the characteristics of the superficial tissue with technical parameters defined by the examiner. In T1 sequence, the main sources of more shinny images are fats, cholesterol crystals, hyperprotein secretions, melanin and acute hemorrhage, being the best sequence to anatomically analyze the region in study. In the T2 sequence, the shiny images are frequently related to presence of fluids (edema, cystic liquids, etc.) 16, 26.

In T1, PA presents an iso-signal in relation to the muscles and in T2, iso and/or hyperintense signal. The increase in intensity of signal in T2 can be attributed to large amounts of mixoid tissue or cystic areas of neoplasm 16. However, the pattern of hyperintensity of signal is also found in lesions that do not present cystic areas and with small amounts of mixoid tissue. The presence of the surrounding line of hypointense signal could be related to capsule image. One of the main limitations of MRI in the diagnosis of PA is the inability of the technique to differentiate the areas of dystrophic calcifications from focal areas of fibrosis 8, 11, 16. The characteristics of different structures comprised by PA are summed up in Chart 2.

MRI has shown many advantages in the diagnosis of PA when compares to other imaging diagnosis, such as CT scan or US. Among them, we can include:

 Multi-plan projections (axial, coronal and sagittal) free from artifacts, obtained without changing the patients' position, with no injection of contrast and ionizing radiation, allow the topographic identification and its relation with the surrounding anatomical structures 9-11;
 The excellent contrast of soft tissues allow differentiation between intra and extraglandular lesions and the best distinction between salivary gland, lesions and adjacent tissues 9-11;
 The intense signal of intermuscular fat in T1 determines an excellent individualization of muscles and low signal of blood fluid allows visualization of larger vessels 10;
 Better definition of the lesion margins, except when there are areas of dystrophic calcifications in which the images by CT scan are superior 11;
 Dynamic studies using MRI have suggested the combination of dynamic curve standards of enhance with contrast and characteristics of the tumor margin in salivary gland lesions can support the differentiation of benign, malignant tumors or inflammatory lesions 12.

The summary of the studied literature (Table 1) highlights 18 studies that totaled the occurrence of 588 cases of pleomorphic adenoma within 27 years (1975 to 2002), being that 431 cases in epidemiological studies 1, 3, 12, 20, 97 cases in clinical trials 4, 5, 9-11, 14, 17, 21-25, 27, and 60 cases of histology studies 13.

In 2 of the 18 studies presented in the table, conventional x-ray was ordered previously to other techniques, being that in both cases the palate was involved 24, 25. The most frequently ordered imaging exam was CT scan, presented in 7 studies. Generally speaking, the main indications for use included definition of limits of the lesion, presence of calcifications inside it, dimensions, precise location, multiple lesions, identification of lesion characteristics and possible bone abnormalities 10, 11, 14, 23-25, 27. Two studies assessed the importance of CT scan in the diagnosis of PA 23, 27. The second test was ordered by MRI, present in 5 studies being that 4 of them were followed up by CT scan 9-12, 27. US was used in 3 studies in the reviewed literature 5, 22, 27.

Sections are frequently essential to plan treatment of pleomorphic adenoma of minor and major salivary glands, since treatment for such damage is normally complete surgical removal. Within such context, studies 7, 10 have demonstrated that MRI has a key role in lesions that involve deep spaces of the head and neck region, because it allows good discrimination between lesion and adjacent tissues, owing to the contrast of hypersignal in T1 of fat in the most central space of the region - the parapharyngeal space. In cases of neoplasm of the parotid region, the importance of preserving the facial nerve has been emphasized, even if it is not involved, being that the analysis is normally made through sections. Incomplete removal of PA can result in recurrence, normally with multiple tumor foci 7, 25. However, authors suggest that each recurrence increases the possibility of malignant transformation 8, 15, 18, 28. CT scan proved to be superior in identifying the presence of calcifications and areas of necrosis in the tumor mass, signals which represent, respectively, benign aspects and possible malignant transformation of the lesion 7, 8, 11, 14, 15, 17. The recurrence varies from 5% to 50%, presenting small growth, which can be diagnosed again 10 to 20 years after the first surgical intervention 27.





CASE REPORT 1

Female patient aged 43 years, referred by the general dental clinician to the discipline of Stomatology FO/UFG, complaining of "lump on the roof of the mouth".

Clinical intraoral examination revealed total upper prosthesis in poor conditions, made more than 20 years before. In the internal surface of the prosthesis, there was the impression of a concavity that coincided with the volume increase on the hard palate on the left, measuring approximately 3.0 x 2.5cm, elastic consistency and asymptomatic upon palpation, with whitish color on the peripheral area and erythema on the central region (Figure 1).

There were no significant bone abnormalities in the panoramic and occlusion x-rays and aspiration puncture was negative.

Microscopic sections of incision biopsy showed epithelial proliferations in nests, cords and ductiform structures with cuboid or flat luminary cells. In some areas, the ductal structures were filled with hyaline amorphous material, recovered by fibrous connective tissue, showing focal areas of dense collagen build-up, amorphous hyaline areas, mixoid areas and chondroid areas, whose anatomical pathologic diagnosis was pleomorphic adenoma.

MRI was conducted right after the incision biopsy, which was not charged from the patient. The images confirmed that it was an oval well-delimited lesion, with regular contours in the left anterior-lateral region of the hard palate. The lesion presented as slightly heterogeneous, with iso-signal to muscle in T1 (Figure 2a) and slightly hyperintense in T2 (Figures 2c, 2d), with moderate enhance after infusion of paramagnetic contrast (Figure 2b). The lesion was approximately 2.1 x 1.4 x 1.2 cm in the largest diameter, without clear cleavage plan with the floor of the nasal cavity and maxillary sinus on the left (Figures 2b, 2d). We also noticed gross thickness of the maxillary sinus mucosa and some ethmoidal cells, whose hyperintense signal in T2 suggested sinusitis of the paranasal sinuses (Figures 2b, 2c and 2d).

Total exeresis of the lesion was made, leaving an oral-sinusal communication that persisted 30 days postoperatively. The patient returned for reassessment only 6 months after surgery, when we detected complete scarring by second intention and absence of oral-sinusal communication. The clinical pathology of the piece confirmed the diagnosis of pleomorphic adenoma (PA).

CASE REPORT 2

Male patient, aged 42 years, leukoderma, was referred from NAPEO - Núcleo de Atendimento ao Paciente Especial em Odontologia (Itumbiara-GO) to Centro Goiano de Doenças da Boca - CGDB/FO/UFG (Goiânia-GO), with diagnostic hypothesis of oropharynx neoplasm with expansion to the soft palate. The patient had already been submitted to CT scan and previous biopsies of the soft palate region, whose clinical pathology analyses were compatible with normal glandular tissue or adenomatoid hyperplasia of drug origin, subject to clinical confirmation.

The clinical examination revealed that the main complaint was a "ball growing in the mouth" for approximately 10 years, and the growth in the oropharynx had increased in the past 3 months. He associated oropharynx anatomy abnormalities with snoring during sleep. The medical history revealed surgery to remove the gallbladder and that the patient was not using any medication. Dental and family history did not produce any significant data. The patient had smoked for 17 years (12 cigarettes/day) and had quit one year before after a medical visit.

Intraoral physical examination showed an increase of well-delimited volume, located on the region of the soft palate on the right, extending to the oropharynx, recovered with slightly red mucosa measuring approximately 3.5cm in the largest diameter, asymptomatic, presenting irregular surface with surgical scar and resilience to palpation (Figure 3).

In the CT scan sent with the patient, we observed an oval lesion, with soft part density, clear contours, no calcifications inside it, causing medial displacement of the parapharyngeal space fat and oropharynx lateral wall (Figure 4a), with slight enhance after injection of contrast (Figure 4b). We also observed opening of the laminae of the right pterygoid process of the sphenoid bone (Figure 4a), suggesting slow growth with bone remodeling, compatible with biological involvement of benign lesions, However, the exam was not helpful to define the limits of the lesion relative to the parotid gland, carotid space and mastication muscles (Figures 4a and 4b).

For differential diagnosis and better visualization of the components of the anatomical spaces we indicated MRI. The images obtained in a magnetic field of 2.0 Teslas, using sequences Spin Echo T1 sagittal, axial, coronal and Fast Spin Echo T2 axial, revealed an expansive oval lesion, with iso-intense signal in T1 relative to muscles (Figures 5a and 5b), heterogeneous hyperintense signal in T2 (Figure 5d), located predominantly in the right parapharyngeal space, without clear cleavage plan with the parotid space (Figure 5a and 5d). The lesion had approximately 5.5 x 4.1 cm in diameter and presented moderate and heterogeneous enhance in T1 after infusion of paramagnetic contrast (Figure 5c). We also observed compression and opening of surrounding anatomical structures, with no signals of invasion and absence of regional adenomegalia (Figures 4 and 5).

The patient was referred to treatment at Hospital Araújo Jorge, Associação de Combate ao Câncer em Goiás. After intraoperative frozen biopsy, which did not reveal signals of malignant transformation, the neoplasm was removed using intraoral surgical route. The clinical pathology analysis of the surgical piece revealed absence of cell atypia, presence of myoepithelial cell proliferation, ductiform structures, mesenchyma components similar to chondroid areas, recovered by a thin fibrous capsule, compatible with PA of parapharyngeal space. Forty days after surgery, the patient progressed well and uneventfully, with normal scarring, and he is currently still being followed up.







DISCUSSION

Imaging diagnosis of pleomorphic adenoma varies a lot depending on the method used. In conventional x-rays, such as in case 1, no significant bone deformities are identified. According to Pinto et al. 21, when the bone destruction of hard palate is observed, it can be associated with the damage, being suggestive of malignancy.

The indication of US in both cases was not the best option, because they were not external lesions and it is quite difficult to have access with the transducers 5-7.

In case 2, after excluding the possibilities of minor salivary gland neoplasm in the region of the soft palate owing to the clinical pathology result, and of medication adenomatosis suggested by the pathologist, considering that the patient did not use any type of medication, the diagnostic hypothesis was damage stemmed the deep structures of the oropharynx. This anatomical location was one of the main justifications for ordering section imaging of the case, which provided analysis of the relation between the neoplasm and the anatomical spaces and their contents.

As to parapharyngeal tumors, Lloyd et al.10 demonstrated that MRI allows good discrimination between the muscle and the lesion, owing to the contrast between the hypersignal in T1 of fat in the region. MRI evidences the displacement of the vessels in the carotid space. In general, in parotid and parapharyngeal space tumors, the carotid artery and the internal jugular vein can be placed concentrically in the carotid space or be displaced posteriorly, differently from neural tumors, in which 2/3 of the cases present anterior displacement of the vessels 10.

All neoplasms of salivary glands are better identified by MRI than by CT scan 7. However, when calcifications and necrosis areas are present in the tumor mass, the CT scan is better 7, 11. In two clinical cases presented here, there were no signs of calcification.

The PA images, in general, both by CT scan and MRI, evidence lesions with benign characteristics, being that larger ones have homogenous appearance whereas large ones have heterogeneous appearance 7.
The literature reveals the consensus that MRI is better than CT scan in the diagnosis of salivary gland lesions 10. In cases in which MRI is accessible, routine or pre-surgical CT scan has been considerable dispensable 18. However, two important aspects have to be considered:

1. The central area of the necrosis in PA that suffer malignant transformation as well as the invasion and destruction of the bone cortical area caused by these lesions are better visualized by CT scan 11. In the modality of image, attenuated areas reduced in the center of the tumor mass have been related to necrosis or cystic abnormalities 7, 11, 29.
2. In external heterogeneous lesions that suggest complex tissue architecture, it is not possible to precisely correlated in which specific cell the lesion provides specific signal. In such cases, MRI does not provide addition information if compared to CT scan 11.

In palate lesions (Case 1), twice smaller than the parapharyngeal space (Case 2), there was predominance of mixoid tissue, whereas in the latter the pattern was more solid (cellular). As to MRI signal, the aspect could correspond to histopathology differences explained by mild heterogeneity of iso-signal in T1 in case 1 (Figure 2a), if compared to the more homogenous iso-signal in the lesion of case 2 (Figure 5a). The intensities of iso-signal in T2 and after infusion of contrast were similar for both cases, despite the larger amount of mixoid areas and ductiform structures with mucoid tissues in case 1, which would suggest a more intense signal in T2 16. Further studies are required to confirm these hypotheses.

The evident connective capsule in case 2, possibly related to the hyposignal line that surrounded the image of the lesion in T1 (Figure 9b) was not identified in the palate lesion (case 1), a lesion that presented less defined limits in MRI, suggesting incomplete encapsulation, as described by the literature 7, 8, 14, 15, 17.

In case 2, MRI was superior to CT scan, since in addition to allowing more precise identification of the lesion limits relative to the anatomical spaces and structures, it also confirmed the presence of a satellite lesion, suggested in post-contrast images of CT scan, ruling out the possibility of lymphadenopathy.

For Case 1, the lack of cleavage plan between the lesion and the floor of the oral cavity, as well as the adjacent left maxillary sinus (Figures 4b, 4d) were essential information, provided by MRI, to support the surgeon in surgical planning, including the prediction of possible buco-naso-sinusal communication postoperatively.

As to treatment, all information provided by the images was essential, not only for surgical planning, but also for differential diagnosis 10, 19.

The benign neoplasms that are more frequent in the parapharyngeal space include lipoma and pleomorphic adenoma, being that the latter is the most prevalent lesion in the parotid space. In carotid space, paraganglioma and neurogenic neoplasm are the most frequent 8, 10, 11, 29, 30. In Case 2, despite the information provided by MRI, it was not possible to define whether lesion was located only in the parapharyngeal space or if it came from the parotid space. Based on such considerations, the clinical hypothesis suggested for the surgical indication was pleomorphic adenoma. The difficulty to access the lesion was essential for the definition of a treatment plan that included intraoperative frozen biopsy followed by enucleation of the lesion.

In Case 1, the option of incision biopsy previously to the lesion excision, was due to the fact that between the most prevalent lesions of the hard palate, there were malignant neoplasms such as adenocarcinoma 3 that can clinically simulate a benign lesion. Moreover, its location was easy to access. The indication of MRI was especially for surgical planning, since the clinical pathology diagnosis had already been defined. The choice of method justified by non-biological invasiveness and superiority to assess soft tissues, associated with the availability of the patient at the moment. The CT scan could have also been the exam of choice, since it was a soft tissue lesion limited by bone tissue.

Therefore, we suggest that if there is availability of MRI, it should be the first option for differential diagnosis and treatment plan of pleomorphic adenoma of salivary glands. The indication of CT scan as the first option would be limited to cases where MRI cannot have access or as a complementary exam, to analyze the invasion of cortical bone layers and/or the presence of calcifications and intra-tumor necrosis.

CONCLUSION

Based on literature review, illustrated by case reports, we could conclude that MRI and CT scan can provide information that are complementary to the diagnosis of pleomorphic adenoma. The main aspects that justify the superiority of MRI to that end are:

 High contrast of soft tissues at multi-plan sections, without changing the patients' position;
 No biological invasion;
 Detailed anatomical and physiological information, characterizing the mixed aspects of the neoplasm and their correlation with anatomical spaces and structures.

REFERENCES

1. Chidzonga MM, Perez VML. e Alavarez ALP. Pleomorphic adenoma of the salivary glands. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1995; 79(6): 747-9.
2. Taylor J, Tighe JVP. A minor salivary gland tumour presenting with dysphagia. J Laryngol Otol 1999 Jun; 113(6): 569-72.
3. Waldron CA, El-Mofty SK, Gnepp DR. Tumours of the intraoral minor salivary glands: A demographic and histologic study of 426 cases. Oral Surg Oral Med Oral Pathol 1988; 66(3): 323-32.
4. Nardone M, Ferrrara G, Nappi O, Di Maria D, Villari G. Pleomorphic adenoma in unusual anatomic sites: case reports and review of literature. Acta Otrhinolaryngol Ital 2002 Jun; 22(3): 158-63.
5. Ishii J, Nagasawa H, Wadamori T, Yamashiro M, Ishikawa H, Yamada T, et al. Ultrasonography in diagnosis of palatal tumours. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999 Jan; 87(1): 39-43.
6. García CJ, Flores PA, Arce JD, Chuaqui B, Schwartz DS. Ultrasonography in the study of salivary gland lesions in the children. Pediatr Radiol 1998; 28(6): 418-25.
7. Silvers AR, Som PM. Salivary Glands. Radiol Clin North Am Head and Neck Imaging 1998 Sep; 36(5): 941-66.
8. Som PM. Salivary Glands. In: Som PM, Bergeron RT. Head and Neck Imaging. 2nd ed. St. Louis: Mosby; 1991. P. 277-348.
9. Austin RJ e Crockett DM. Pleomorphic adenoma of the palate in child. Head & Neck 1992 Jan-Feb; 14(1):58-61.
10. Lloyd GAS, Phelps PD. The demonstration of tumors of the parapharyngeal space by magnetic ressonance imaging. Br J Radiol 1986; 59(703): 675-83.
11. Som PM, Shugar JMA, Sacher M, Stollman AL, Biller HF. Benign and Malignant Parotid Pleomorphic Adenomas: CT and MR Studies. J Comput Assist Tomogr 1988; 12(1): 65-9.
12. Suenaga S, Indo H, Noikura T. Diagnostic value of dynamic magnetic resonance imaging for salivary gland diseases: a preliminary study. Dentomaxillofac Radiol 2001; 30:314-8.
13. Alves FA, Perez DE, Almeida OP, Lopes MA, Kowalski LP. Pleomorphic adenoma of submandibular gland: clinicopathological and immunohistochemial feateure of 60 cases in Brazil. Arch Otolaryngol Head Neck Surg 2002 Dec; 128(12): 1400-3.
14. Erol B, Özer N, Unlü G, Gülsün B. A massive pleomorphic adenoma of the submandibular salivary gland. J Oral Maxillofac Surg 1997; 55: 1011-4.
15. Neville BW, Damm DD, Allen CM, Bouquot JE. Patologia das Glândulas Salivares. In: __ Patologia Oral & Maxillofacial. Rio de Janeiro: Guanabara Koogan; 1998. p. 314-52.
16. Soller R, Bargiela A, Requejo I, Rodríguez E, Rey, JL, Sancristan, F. Pictorial review: MR Imaging of Parotid Tumors. Clin Radiol 1997; 52(4): 269-75.
17. Takeda Y, Yamamoto H. Stromal Bone Formation in Pleomorphic Adenoma of Minor Salivary Gland Origin. J Nihon Univ Sch. Dent 1996; 38(2): 102-4.
18. Regezi JÁ, Sciubba JJ. Salivary Glands. In: __ Oral Pathology: Clinical pathologic correlations. 3rd ed. Philadelfia: Saunders; 1999. p.195-243.
19. Touquet R, Mackenzie J, Carrut JAS. Management of the Parotid Pleomorphic Adenoma, the problem of exposing tumour tissue at operation. The logical pursuit of treatment policies. Br J Oral Maxillofac Surg 1990; 28(6): 404-7.
20. Landini G. Nucleolar Organizing Regions (NORs) in pleomorphic adenomas of the salivary glands. J Oral Pathol Med 1990; 19: 257-60.
21. Pinto RS, Kelly DE, George AE. Radiologic features of benign pleomorphic adenoma of the hard palate. Oral Surg Oral Med Oral Pathol 1975 Jun; 39(6): 976-81.
22. Shimizu M, Ussmüller J, Hartwein J, Donath K, Kinukawa N. Statistical study for sonographic differential diagnosis of tumorous lesions in the parotid gland. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999 Aug; 88(2):226-33.
23. Kurabayashi T, Ida M, Yoshino N, Sasaki, Ishii J, Ueda M. Differential diagnosis of tumours of the minor salivary glands of the palate by computed tomography. Dentomaxillofac Radiol 1997; 26: 16-21.
24. Boneu F, González-Lagunas J, Huguet P, Bassas C. Massive Malignant Pleomorphic Adenoma of the Palate. J Oral Maxillofac Surg 1998; 56: 91-6.
25. Goodisson DW, Burr RGM, Creedon AJ, Stirling RW, Morgan PR, Odell, EW. A Case of metastasing pleomorphic adenoma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999 Mar; 87(3):341-5.
26. Langlais RP, van Rensburg LJ, Guidry J, Moore WS, Miles DA, Nortjé CJ. Magnetic resonance imaging in dentistry. Dent Clin North Am 2000 Apr; 22(2): 411-26.
27. Yasumoto M, Sunaba K, Shibuya H, Kurabayashi T. Recurrent pleomophic adenoma of the head and neck. Neuroradiology 1999 Apr; 41(4): 300-4.
28. Shafer WG, Hine MK, Levy BM. Tumores das Glândulas Salivares. In: __ Tratado de Patologia Bucal. 4a. ed. Rio de Janeiro: Guanabara Koogan; 1987. p. 213-38.
29. Som PM. Parapharyngeal Space. In: Som PM, Bergeron RT. Head and Neck Imaging. 2nd ed. St. Louis: Mosby; 1991. p. 467-96.
30. Delbalso AM, Ellis GE, Hartman KS, Langlais RP. Diagnostic Imaging of the Salivary Glands and Periglandular Regions. In: Delbalso AM. Maxillofacial imaging. Philadelphia: Saunders, 1990. p. 409-510.
31. Vogl TJ, Balzer J, Mack M, Steger W. Parapharyngeal Space and Adjacent Deep Compartments. In: __Differential Diagnosis in Head and neck Imaging: A systematic approach to the radiologic evaluation of the head and neck interpretation of difficult case. Stuttgart: Thieme; 1999. p. 215-36.

---

1 Joint Professor, Discipline of Semiology and Stomatology, Department of Stomatology Sciences, FO/UFG, Coordinator of Centro Goiano de
Doenças da Boca/FO/UFG, Visiting Researcher, Eastman Dental Center, University of Rochester, New York, USA.
2 Undergraduate, School of Dental Sciences, Trainee at Centro Goiano de Doenças da Boca, FO/UFG.
3 Master studies under course, Program of Master Studies in Bucomaxillofacial Radiology, FO/UFG; Assistant Professor, Discipline of Radiology, FO Anápolis.
4 Faculty Professor, Discipline of Oral Pathology, Department of Stomatology Sciences, FO/UFG, Post-doctorate, University of Texas at San Antonio, USA.
5 Dental surgeon, Specialist in Bucomaxillofacial Surgery and Traumatology and Oral Pathology.
6 Radiologist, Fundação José Normanha, Instituto Goiano de Radiologia, Goiânia-GO.
School of Dental Sciences - Federal University of Goiás - Centro Goiano de Doenças da Boca (CGDB)
Address correspondence to: Rejane Faria Ribeiro-Rotta - Rua C235 nº 1323 ap. 1501 Edifício Leblon Setor Nova Suíça Goiânia GO 74280-130
Tel (55 62) 259-3766 - E-mail: rrrotta@odonto.ufg.br
Study presented at 12º Congresso Internacional de Odontologia de Goiás, September 2002.
Article submitted on February 11, 2003. Article accepted on May 13, 2003.