Real-Time Sonoelastography of Major Salivary Gland Tumors

Transcription

1 Neuroradiology/Head and Neck Imaging Original Research Dumitriu et al. Sonoelastography of Major Salivary Gland Tumors Neuroradiology/Head and Neck Imaging Original Research Dana Dumitriu 1 Sorin Dudea 1 Carolina otar-jid 1 Mihaela ăciuț 2 Grigore ăciuț 2 Dumitriu D, Dudea S, otar-jid C, ăciuț M, ăciuț G Keywords: neoplasm, salivary gland, sonoelastography DOI: /JR Received January 18, 2011; accepted without revision pril 8, Department of Radiology, University of Medicine and Pharmacy Iuliu Hatieganu Cluj-Napoca, 1 3 Clinicilor St, Cluj-Napoca , Romania. ddress correspondence to D. Dumitriu (danamrc@gmail.com). 2 Department of Oral and Maxillofacial Surgery, University of Medicine and Pharmacy Iuliu Hatieganu Cluj-Napoca, Cluj-Napoca, Romania. WE This is a Web exclusive article. JR 2011; 197:W924 W X/11/1975 W924 merican Roentgen Ray Society Real-Time Sonoelastography of Major Salivary Gland Tumors OJECTIVE. The purpose of this study was to determine the performance of real-time sonoelastography in the differential diagnosis of salivary gland tumors. SUJECTS ND METHODS. etween 2007 and 2010, 74 salivary gland tumors were examined by ultrasound and sonoelastography in 66 patients. Lesions were graded according to a 4-point elastography score. Surgical excision and histopathologic examination were performed in all cases. The difference in elastographic score between benign and malignant masses and that between pleomorphic adenomas and Warthin tumors were evaluated. RESULTS. Of the 74 salivary tumors, 63 were located in the parotid, and 11 were in the submandibular gland. There were 18 malignant and 56 benign tumors. The mean (± SD) elastographic score was 2.58 ± 0.87 for pleomorphic adenomas, 2.15 ± 0.80 for Warthin tumors, 2.00 ± 0.57 for other benign tumors, and 2.94 ± 0.87 for malignant tumors. For benign tumors overall, the mean elastographic score was 2.41 ± The difference in elastographic score between benign and malignant tumors overall was statistically significant (p < 0.05), but the difference between malignant tumors and pleomorphic adenomas and that between Warthin tumors and pleomorphic adenomas were not statistically significant. Using cutoff values between scores 2 and 3 and scores 3 and 4, there was no statistically significant difference between benign and malignant tumors. CONCLUSION. lthough this study revealed a difference in elastographic score between benign and malignant tumors, detailed analysis did not provide consistent results. Consequently, real-time sonoelastography appears to be a limited technique in the differential diagnosis between benign and malignant salivary masses. T he observation that malignant lesions are generally stiffer than benign ones has made real-time sonoelastography an interesting option in the differential diagnosis of focal masses in various organs, the most studied to date being breast, thyroid, and prostate lesions. ecause salivary gland masses are mostly superficial and because it is notoriously difficult to determine the pathologic type by use of any imaging technique [1 12], the question has arisen whether elastography might bring additional information to the differential diagnosis and thus facilitate the choice of surgical technique. The main question for the surgeon is whether a salivary mass is benign or malignant, whereas differentiating between the different tumor types, especially the most frequent benign masses (i.e., pleomorphic adenoma and Warthin tumor), is of secondary importance, because the surgical technique will not differ substantially [13]. The aim of the current study was to determine whether using an elastography score similar to the one used for breast or thyroid nodules might help to differentiate between benign and malignant masses of the parotid and submandibular glands. Subjects and Methods Patients etween pril 2007 and September 2010, 74 salivary gland masses were examined in 66 consecutive patients. Twenty-nine patients were female and 37 were male (mean [± SD] age, 50.8 ± 2.07 years; range, 6 82 years). Sixty-three masses (85.1%) were located in the parotid, and 11 (14.9%) were in the submandibular gland. Patients were included in the study if they presented with a palpable mass that was located within a salivary gland on ultrasound. Patients were excluded if ultrasound revealed a mass that was in the parotid or submandibular spaces but clearly outside the gland. lso, images of isolated intraparotid W924 JR:197, November 2011

2 Sonoelastography of Major Salivary Gland Tumors lymph nodes of normal shape and size were excluded from analysis. The study was approved by the ethics committee of the institution. Informed consent was obtained from all patients. TLE 1: Description of Sonoelastography Scores Score Description 1 Mass is similar in elasticity to surrounding glandular parenchyma, with a mixture of green, yellow, and red areas (Fig. 1) 2 Mass is predominantly soft compared with adjacent parenchyma, with some areas of stiffness present, representing < 50% of tumor area (Fig. 1) 3 Mass is predominantly stiff, but areas of elasticity are still present; stiffness is present over > 50% of tumor area (Fig. 1C) 4 Mass is entirely stiff (ranging from light blue to dark blue on elastogram) (Fig. 1D) Note The proportion of stiff versus elastic areas in scores 2 and 3 was appreciated subjectively by the examiner. Imaging ll patients were examined by gray-scale and Doppler ultrasound, followed by real-time elastography. oth ultrasound and elastography were performed on an ultrasound machine with a MHz 5-cm-aperture linear probe (EU 8500, Hitachi). The examination protocol consisted of gray-scale ultrasound with measurement of the tumor size and of the distance between the transducer and the external surface of the mass, as well as analysis of the structure of the mass, with a special emphasis on fluid areas. Elastography was performed for each patient according to a standard protocol developed in our department, as follows: the transducer was applied strictly perpendicular to the skin, without the use of a stabilizer; the largest sonoelastographic box possible was used in every case; depth settings were determined by the size of the tumor; transverse and sagittal scans over the tumor mass and one video sequence of at least 10 seconds for each case were recorded; and a compression quality factor between 3 and 5, as indicated by the machine scale, was always used when recording the images. oth tumor and normal gland tissue were included in the elastographic box in all cases where this was possible. The video sequences and static images in both planes were reviewed by one examiner who was blinded to the pathologic findings. 4-point elastographic score adapted from the breast elastography score described by Itoh et al. [14] was used to classify each tumor. The description of the elastographic aspects for each score is provided in Table 1 and is illustrated in Figure 1. The presence of fluid areas was noted, and only the solid portions of the mass were considered when determining the elastographic score. Surgical excision was performed for all of the patients, and the elastographic score was compared with the histopathologic result after analysis of the entire pathologic specimen. Statistical analysis was performed using R software (version , The R Foundation for Statistical Computing). The nonparametric Mann- Whitney U test was used to compare the maximum size for benign and malignant tumors, as well as the maximum size for tumors with or without internal fluid areas. The frequency of internal fluid areas between the main types of tumors was evaluated using the Fisher exact test. We used the Mann-Whitney U test to compare elastographic scores in the group of benign tumors (including pleomorphic adenomas, Warthin tumors, and other benign masses) and malignant tumors (including primary and secondary malignancies). The same test was also used to compare elastographic scores for pleomorphic adenomas versus malignant tumors and Warthin tumors, respectively. C D Fig. 1 Color (left panels) and gray-scale (right panels) examples of sonoelastography scores., Score of 1. rrows denote hypoechoic homogeneous salivary mass, which appears entirely soft (dark green, with some red and yellow) on elastogram., Score of 2. rrows denote well-defined salivary mass, predominantly soft on the color elastogram, with some stiff (blue) areas, representing less than 50% of tumor. C, Score of 3. rrows denote lobulated hypoechoic salivary mass, with a mixed elastography pattern in which stiff (blue) areas represent more than 50% or tumor. D, Score of 4. rrows denote lobulated hypoechoic mass, which is almost entirely stiff (dark and light blue). JR:197, November 2011 W925

3 Dumitriu et al. TLE 2: Ultrasound and Sonoelastographic Features of Parotid and Submandibular Tumors No. of Nodules/ Location, No. (%) of Tumors Presence of Fluid Content, Tumor Type No. of Patients Mean Size (mm) Parotid Submandibular No. (%) of Tumors Mean Elastographic Score Pleomorphic adenomas 36/ (88.8) 4 (11.1) 9 (25) 2.58 Warthin tumors 13/ (100) 0 10 (76.9) 2.15 Other benign tumors a 7/ (85.7) 1 (14.2) 2 (28.5) 2.00 Malignant tumors b 18/ (66.6) 6 (33.3) 6 (33.3) 2.94 a Other benign tumors include myoepithelioma (n = 1), oncocytoma (n = 1), lipoma (n = 1), neurofibroma (n = 1), benign lymphoepithelial lesion (n = 1), and epidermoid cyst (n = 2). b Malignant tumors include 15 primary tumors mucoepidermoid carcinoma (n = 5), squamous cell carcinoma (n = 3), adenocarcinoma (n = 2), adenoid cystic carcinoma (n = 2), adenosquamous carcinoma (n = 2), and carcinosarcoma (n = 1) and three secondary tumors. The data on elastographic scores were dichotomized using, first, a cutoff value between 2 and 3, to distinguish between benign and malignant tumors, and then a cutoff value between 3 and 4. cutoff value between 2 and 3 to distinguish between Warthin tumors and pleomorphic adenomas was also evaluated, using the Fisher exact test. Dichotomization was also used for tumor size, with a cutoff value of 40 mm between the two groups. The Fisher test was used to evaluate the difference in elastographic score depending on the size category. For all statistical tests used, a p value less than 0.05 was considered statistically significant. Results mong the 74 salivary gland masses, 18 (24.3%) were malignant. The rate of malignancy was higher in masses located in the submandibular gland: six of 11 masses (54.5%) were malignant. For parotid masses, only 12 of 63 (19.04%) were malignant. The most common diagnosis overall was pleomorphic adenoma (36 masses [48.6%] in 32 patients). For three patients, the pleomorphic adenomas represented recurrences after surgical excision; for two of these patients, the recurrences were multinodular, with three identifiable nodules each, whereas for the third patient, the recurrence was solitary. In the malignant tumors group, 15 tumors were primary salivary malignancies, whereas in three cases, the masses represented solitary lymphonodular metastases from other sites. The final histopathologic diagnoses for all cases are provided in Table 2. The mean maximum tumor size for the entire group was ± mm (range, mm). The mean maximum size for malignant tumors was ± mm, and that for benign tumors was ± mm. The Mann-Whitney U test revealed no statistically significant difference between the mean maximum size for benign and malignant tumors (p > 0.05). Fig year-old man with pleomorphic adenoma of parotid. Tumor cannot be completely filled with color on elastogram because of its large size and distance from transducer. Twenty-seven masses (36.48%) presented internal fluid areas. mong these, 10 were Warthin tumors, nine were pleomorphic adenomas, and six were malignant tumors. The remaining two masses with fluid content were, respectively, an epidermoid cyst and a benign lymphoepithelial lesion. The average maximum size for masses with internal fluid areas was ± mm, whereas that for entirely solid tumors was ± mm (p < 0.05). The Fisher exact test did not reveal a statistically significant difference between the incidence of internal fluid areas in benign and malignant tumors; however, within the benign group, the frequency of internal fluid areas was significantly higher in Warthin tumors than in pleomorphic adenomas (p < 0.01). The mean distance between the surface of the tumor and the transducer was 4.41 ± 0.25 mm (range, 1 12 mm). In three cases, the tumors could not be entirely filled with color on the elastogram. In these cases, the maximum size was 42, 20, 69.8 mm, respectively, and the maximum distance from the transducer was 8, 9, and 10 mm, respectively (Fig. 2). The distribution of elastographic scores for each tumor category is presented in Table 3. The mean elastographic score was 2.58 ± 0.87 for pleomorphic adenomas, 2.15 ± 0.80 for Warthin tumors, 2.00 ± 0.57 for other benign tumors, and 2.94 ± 0.87 for malignant tumors. For benign tumors overall (pleomorphic adenomas, Warthin tumors, and other benign tumors), the mean elastographic score was 2.41 ± The Mann-Whitney U test for the elastographic scores of benign versus malignant tumors revealed a statistically significant difference (p = 0.02). The difference in elastographic score was not statistically significant between pleomorphic adenomas and malignant tumors (p = 0.14); the difference in elastographic score was also not statistically TLE 3: Distribution of Elastographic Scores by Tumor Type Tumor Type Score 1 Score 2 Score 3 Score 4 Pleomorphic adenoma (n = 36) 3 (8.3) 15 (41.6) 12 (33.3) 6 (16.6) Warthin tumor (n = 13) 3 (23.0) 5 (38.4) 5 (38.4) 0 (0) Other benign tumors (n = 7) 1 (14.2) 5 (71.4) 1 (14.2) 0 (0) Malignant tumors (n = 18) 1 (5.5) 4 (22.2) 8 (44.4) 5 (27.7) Note Data are no. (%) of tumors. W926 JR:197, November 2011

4 Sonoelastography of Major Salivary Gland Tumors Fig. 3 Pleomorphic adenomas., 63-year-old man with pleomorphic adenoma of parotid (score 2)., 19-year-old man with pleomorphic adenoma of submandibular gland (score 4). significant between pleomorphic adenomas and Warthin tumors (p = 0.17). To assess the ability of the proposed elastographic score to differentiate between benign and malignant lesions and between pleomorphic adenomas and Warthin tumors, contingency tables were used. Dichotomization was set successively at cutoff values between 2 and 3 and between 3 and 4. The Fisher exact test revealed no statistically significant difference, in either case. The Fisher exact test did not reveal a significant difference in the distribution of elastographic scores between tumors smaller than 40 mm and those 40 mm or larger in maximum size. Discussion Salivary gland tumors form a heterogeneous group of masses, which may be located in both the major and minor salivary glands [15]. particular difficulty when studying these tumors is the large number of pathologic entities, with certain types of benign or malignant tumors being exceedingly rare. It is therefore difficult to identify specific features for each type of mass. However, most tumors of the major salivary glands are benign pleomorphic adenomas; despite their benign nature, they may recur and become malignant in rare cases [16]. Furthermore, the histologic pleomorphism often reflects an imaging pleomorphism. Warthin tumors are the second most common type of benign mass [15]. They rarely occur outside major salivary glands and may be bilateral and cystic [4]. Malignant tumors represent less than 20% of parotid masses, but they are more common in the submandibular, sublingual, and minor salivary glands; their behavior, as well as their imaging features, ranges from pseudobenign to very aggressive [15]. It has therefore been stated that a salivary tumor s behavior depends less on the histologic type and more on the size of the mass, with some researchers considering 4 cm as a cutoff point [17]. Many features of salivary masses have been evaluated in the hope of finding the ideal one to distinguish benign from malignant tumors, because this is the main point of interest before surgery. Our aim was to determine whether examining another feature of these masses, their elasticity, might help in the differential diagnosis. Real-time sonoelastography is a simple technique, used at present mostly in breast and thyroid pathology [14, 18, 19]. n elastography score was first developed for breast lesions [14] and later was adapted for thyroid masses in various ways, using either a 4-point [18] or a 6-point scale [19]. We used a 4-point scale in our study, which is detailed in Table 1; globally, types 1 and 2 on this scale represent lesions with predominant strain, whereas types 3 and 4 include those lesions with little or no strain. The distribution of masses in our group of patients was influenced by several factors: first, all of the patients in our group presented with either parotid or submandibular masses, with no sublingual or minor salivary masses presenting during the period of recruitment. This obviously reflects the incidence of malignant masses, because they are more frequent in minor salivary glands [15]. Most masses were pleomorphic adenomas, with malignant tumors coming second and Warthin tumors third. Seven masses were of other benign histologic types. This distribution reflects the one generally reported for the histology of parotid and submandibular tumors [15]. mong malignant tumors, the most common was mucoepidermoid carcinoma; this finding is also in agreement with commonly reported incidences for malignant salivary tumors [17]. We may therefore infer that the distribution of tumors in our group of patients is similar to the general prevalence of these masses, as presented in the literature. We previously reported the preliminary results of this study, stating that there might be a difference of elasticity between benign and malignant masses (Dumitriu D, et al., presented at the 2008 annual meeting of the Radiological Society of North merica) and that a difference of pattern between pleomorphic adenomas and Warthin tumors should be taken into consideration (Dumitriu D, presented at the 2008 European Congress of Ultrasound Euroson). lthough, overall, we did find a significant statistical difference between the elastographic scores of malignant and benign tumors, when taking into account the difference between pleomorphic adenomas and malignant tumors, we found no significant difference. This is likely because pleomorphic adenomas have an important fibrous component, which increases tumor stiffness (Fig. 3). ecause overall more than half of salivary gland tumors and as many as 80% of parotid gland tumors are pleomorphic adenomas [15], it becomes clear that other benign types are sporadic. Even though rare entities such as oncocytoma and myoepithelioma were present in our group, a specific feature could not be identified for them because of their reduced number. Thus, it is essential to be able to distinguish pleomorphic adenomas from malignant tumors, and we found no statistically significant difference between the elastography scores of the two groups. Warthin tumors are generally well-defined masses, with a regular contour, internal fluid areas, and hypervascularization (Fig. 4). However, pleomorphic adenomas may also be quite well vascularized and contain fluid areas (Fig. 5). To date, most studies agree that the most distinctive feature of pleomor- JR:197, November 2011 W927

5 Dumitriu et al. Fig. 4 Warthin tumors., 67-year-old man with Warthin tumor of parotid. Mass is solid and with almost no strain (score 2)., 60-year-old woman with Warthin tumor of parotid. Solid content (score 3) and large fluid areas (arrow) are seen, with blue-green-red color stratification. Fig. 5 spect of fluid content in tumors., 51-year-old man with parotid pleomorphic adenoma. Central fluid area with blue-green-red color stratification (arrow) is seen on elastogram., 47-year-old man with Warthin tumor of parotid. Tumor is almost entirely fluid, with less regular color stratification, and there was difficulty in determining elastographic score for deep solid segment. phic adenomas on ultrasound is their lobulated contour [2, 4, 20]. On the other hand, distinguishing between the two masses is not crucial because the surgical approach is similar for the two [13]. The most important imaging diagnostic distinction is between benign and malignant masses for the selection of appropriate therapy. lthough an isolated parotidectomy or submandibulectomy is reserved for benign tumors, in case of a malignant tumor, radical parotidectomy or supraomohyoid dissection and simultaneous neck dissection according to the N factor in the TNM staging are indicated. lso, the surgical algorithm for the facial nerve and lingual nerve is selected accordingly [21]. In this study, the mean elastographic score for pleomorphic adenomas was 2.58, whereas that for Warthin tumors was We found no statistically significant difference between them. This is different from the result obtained by hatia et al. [22]. The explanation might lie in several elements, including the difference in fluid content between their group and ours, the higher proportion overall of Warthin tumors in their group (where Warthin tumor was the most frequent mass), and the difference in elastography technique and manufacturer. For Warthin tumors, we encountered a technical problem: 10 of the 13 (76.9%) Warthin tumors in our group presented liquid areas inside, one of them being almost completely fluid, with only a deep solid segment (Fig. 5). This is very common for this type of tumor, and the higher incidence of fluid content reported is no different from what is commonly described in the literature [4, 5, 16]. However, our observation suggests that the compression of the solid part may be affected by the lack of strain of the fluid part, giving an aspect of no strain in the direction of the axis of compression. lthough there is no means of statistically verifying this supposition in the current study, it has also arisen from the observations of hatia et al. [22]. This represents an important problem for elastography: the liquid areas either will be completely strain free and black on the elastogram or will present a color-stratification pattern, as described by Itoh et al. [14] for breast lesions. lthough the blue-greenred pattern is the classic one described, other stratification patterns are possible and have been described in large breast lesions [23]; we also found them in our group of tumors. Most malignant masses were either score 3 or 4 (Figs. 6 and 6C), but among tumors with a score of 4, more than half (6/11) were pleomorphic adenomas. lso, five malignant masses were either score 1 or 2 (Table 3 and Figs. 6 and 6D). ecause of the small number of malignant tumors in this group, a statistical comparison between the subtypes of malignancy is not possible. The mean elastographic score for malignant tumors was 2.94 ± The difference from pleomorphic adenomas was not statistically significant, and so we must conclude that features other than elasticity should be used for the differential diagnosis between these two types of tumors before surgery. Overall size did not influence the distribution of elastographic scores. The difference in elastographic score between the tumors that were smaller than 40 mm and those 40 mm or W928 JR:197, November 2011

6 Sonoelastography of Major Salivary Gland Tumors Fig. 6 Elastographic aspects of malignant salivary masses., 58-year-old woman with mucoepidermoid carcinoma of parotid (score 3)., 52-year-old man with metastatic parotid lymph node from malignant melanoma (score 2). C, 65-year-old man with adenocarcinoma in anterior segment of parotid (score 4). D, 78-year-old woman with adenoid cystic carcinoma of parotid (score 2). C larger was not statistically significant. It may be inferred that size is not necessarily a factor that influences the elastographic score; however, our observation is that larger masses are more difficult to examine correctly. s revealed by previous articles on sonoelastography technique, it is important when performing the examination to have both tumor and normal tissue in the elastography box, to have as large a box as possible, and to apply gentle perpendicular compression, within the correct range as identified on the machine scale [14, 24]. Having both tumor and normal tissue in the elastographic box can be a challenge in the case of large masses. Technical factors are important when performing elastography. We found that superficial and protruding masses were difficult to examine correctly because it was a real challenge to adapt the transducer completely to the skin. The distance from the transducer may be a hindrance for elastography; this is particularly true for masses located in the deep lobe of the parotid gland, which may be more difficult to examine. The only other study, to our knowledge, published on salivary mass elastography specifically indicates excluding masses located in the deep lobe of the parotid gland [22]. Deep parotid masses were included in the present study. Three situations where the tumor could not be examined elastographically occurred in our group. Two of the tumors in question were situated in the deep lobe of the parotid, at 8 and 9 mm from the surface, respectively, whereas the third tumor was a submandibular mass, which was not only situated at 10 mm from the surface but also very large (69.8 mm). The relatively low number of malignant masses in our group might represent a limitation of the current study, but it reflects the common distribution of salivary tumors. Pathologically, the malignant group was heterogeneous, because several types of primary and secondary tumors were present. Compared with previously published studies [22], one of the strong points of the current study is the fact that each case was confirmed with the pathology report, which was delivered after examination of the entire pathologic specimen. ecause fine-needle aspiration has a lower accuracy for malignant masses in particular [25 28], having the pathologic confirmation for each case adds weight to the diagnoses in this group. In conclusion, although there was a significant difference in elastographic score between benign and malignant masses overall in this group, detailed analysis did not reveal a clear cutoff point between elastography scores. The high degree of stiffness of the most common benign salivary tumor, pleomorphic adenoma, makes elastography a disappointing technique at present for the differential diagnosis between benign and malignant salivary masses. We therefore think that the precise differential diagnosis of salivary masses is still very much a matter of pathology and less of imaging and that, despite promising initial results, qualitative real-time sonoelastography does not seem to have a place in the diagnostic algorithm of salivary gland tumors for the time being. References 1. Schick S, Steiner E, Gahleitner, et al. Differentiation of benign and malignant tumors of the parotid gland: value of pulsed Doppler and color Doppler D JR:197, November 2011 W929

7 Dumitriu et al. sonography. Eur Radiol 1998; 8: Radiology 2005; 237: Shah P, Patel SG. Salivary glands. In: Shah JP, 2. iałek E, Jakubowski W, Karpińska G. Role of 11. Takashima S, Wang J, Takayama F, et al. Parotid Patel SG. Head and neck surgery and oncology. ultrasonography in diagnosis and differentiation masses: prediction of malignancy using magneti- Edinburgh, UK: Mosby Elsevier Limited, of pleomorphic adenomas. rch Otolaryngol zation transfer and MR imaging findings. JR 2003: Head Neck Surg 2003; 129: ; 176: hatia KS, Rasalkar DD, Lee YP, et al. Evalua- 3. Steinhart H, Zenk J, Sprang K, ozzato, Iro H. 12. Fischer T, Paschen CF, Slowinski T, et al. Differ- tion of real-time qualitative sonoelastography of Contrast-enhanced color Doppler sonography of parotid gland tumors. Eur rch Otorhinolaryngol 2003; 260: iałek E, Jakubowski W, Zajkowski P, Szopinski K, Osmolski. US of the major salivary glands: anatomy and spatial relationships, pathologic conditions, and pitfalls. RadioGraphics 2006; 26: Thoeny HC. Imaging of salivary gland tumors. Cancer Imaging 2007; 7: Martinoli C, Derchi L, Solbiati L, Rizzato G, Silvestri E, Giannoni M. Color Doppler sonography of salivary glands. JR 1994; 163: Choi DS, Na DG, yun HS, et al. Salivary gland tumors: evaluation with two-phase helical CT. Radiology 2000; 214: Yabuuchi H, Fukuya T, Tajima T, Hachitanda Y, Tomita K, Koga M. Salivary gland tumors: diagnostic value of gadolinium-enhanced dynamic MR imaging with histopathologic correlation. Radiology 2003; 226: Yabuuchi H, Matsuo Y, Kamitani T, et al. Parotid gland tumors: can addition of diffusion-weighted MR imaging to dynamic contrast-enhanced MR imaging improve diagnostic accuracy in characterization? Radiology 2008; 249: King D, Yeung DK, huja T, et al. Salivary gland tumors at in vivo proton MR spectroscopy. entiation of parotid gland tumors with contrastenhanced ultrasound. Rofo 2010; 182: O rien CJ. Current management of benign parotid tumors: the role of limited superficial parotidectomy. Head Neck 2003; 25: Itoh, Ueno E, Tohno E, et al. reast disease: clinical application of US elastography for diagnosis. Radiology 2006; 239: Pinkston J, Cole P. Incidence rates of salivary gland tumors: results from a population-based study. Otolaryngol Head Neck Surg 1999; 120: Lee YYP, Wong KT, King D, huja T. Imaging of salivary gland tumors. Eur J Radiol 2008; 66: Speight PM, arrett W. Salivary gland tumors. Oral Dis 2002; 8: steria C, Giovanardi, Pizzocaro, et al. USelastography in the differential diagnosis of benign and malignant thyroid nodules. Thyroid 2008; 18: Hong Y, Liu X, Li Z, Zhang X, Chen M, Luo Z. Real-time ultrasound elastography in the differential diagnosis of benign and malignant thyroid nodules. J Ultrasound Med 2009; 28: Dumitriu D, Dudea SM, otar-jid C, ăciuţ G. Ultrasonographic and sonoelastographic features of pleomorphic adenomas of the salivary glands. Med Ultrasound 2010; 12: focal lesions in the parotid and submandibular glands: applications and limitations. Eur Radiol 2010; 20: Chiorean R, Duma MM, Dudea SM, et al. Typical and unusual sonoelastographic patterns of breast cystic lesions: impact on I-RDS classification. Ultraschall Med 2010 Oct 22 [Epub ahead of print] 24. Ciurea, olboaca S, Ciortea C, otar-jid C, Dudea SM. The influence of technical factors on sonoelastographic assessment of solid breast nodules. Ultraschall Med 2011; 32(suppl 1):S27 S Das DK, Petkar M, l-mane NM, Sheikh Z, Mallik MK, nim JT. Role of fine needle aspiration cytology in the diagnosis of swellings in the salivary gland regions: a study of 712 cases. Med Princ Pract 2004; 13: Stanley MW. Selected problems in fine needle aspiration of head and neck masses. Mod Pathol 2002; 15: Howlett DC, Menezes LJ, Lewis K, Moody, Violaris N, Williams MD. Sonographically guided core biopsy of a parotid mass. JR 2007; 188: Handa U, Dhingra N, Chopra R, Mohan H. Pleomorphic adenoma: cytologic variations and potential diagnostic pitfalls. Diagn Cytopathol 2009; 37:11 15 W930 JR:197, November 2011