Sonographically Guided Core Biopsy of A Parotid Mass

Transcription

1 Howlett et al. Sonographically Guided Biopsy of Parotid Mass Head and Neck Imaging Original Research 223.fm 11/30/06 David C. Howlett 1 Leon J. Menezes 2 Khari Lewis 1 Andrew B. Moody 2 Nick Violaris 3 Michael D. Williams 2 Howlett DC, Menezes LJ, Lewis K, Moody AB, Violaris N, Williams MD Keywords: head and neck imaging, head and neck radiology, parotid gland, sonography DOI: /AJR Received September 1, 2005; accepted after revision December 7, Department of Radiology, Eastbourne District General Hospital, East Sussex BN21 2UD, United Kingdom. 2 Department of Maxillofacial Surgery, Eastbourne District General Hospital, King s Dr., Eastbourne, East Sussex BN21 2UD, United Kingdom. Address correspondence to D. C. Howlett (David.Howlett@ESHT.NHS.UK). 3 Department of Ear, Nose and Throat Surgery, Eastbourne District General Hospital, Eastbourne, East Sussex, BN21 2UD, United Kingdom. AJR 2007; 188: X/07/ American Roentgen Ray Society Sonographically Guided Core Biopsy of A Parotid Mass OBJECTIVE. The purpose of this study was to evaluate the accuracy of sonographically guided core biopsy in the evaluation of parotid masses. SUBJECTS AND METHODS. Between 1998 and 2004, 135 patients consecutively presenting with a parotid mass were prospectively enrolled into this study. A single operator performed initial diagnostic sonography and then sonographically guided core biopsy using local anesthesia. Biopsy was performed with an 18- or 20-gauge needle and a spring-loaded biopsy gun with a mean of two passes per patient. Outcome measures were accuracy, sensitivity, specificity, and predictive values of sonographically guided core biopsy compared with the final pathologic diagnosis in the surgical group. In the nonsurgical group, final diagnosis was established on the basis of histologic findings after adequate core biopsy and clinical follow-up. RESULTS. All sonographically guided core biopsy specimens were considered satisfactory for histologic evaluation. Overall there were 71 benign tumors, 35 malignant tumors, and 29 miscellaneous, nonneoplastic lesions. In 76 (56%) of the 135 patients who underwent surgery, sonographically guided core biopsy and surgical histologic findings were correlated for 74 patients. In two cases sonographically guided core biopsy and surgical histologic findings did not correlate. In one case, the sonographically guided core biopsy finding was mucoepidermoid carcinoma, but the final diagnosis was squamous cell carcinoma. In the other case, the finding at sonographically guided core biopsy was squamous cell carcinoma, but the final diagnosis was mucoepidermoid carcinoma. The treatment of these patients was not affected. Fifty-nine (44%) of the 135 patients avoided surgery. In differentiation of benign from malignant disease, sonographically guided core biopsy had a sensitivity, specificity, and diagnostic accuracy of 100%. Sonographically guided core biopsy also had positive and negative predictive values of 100% in the diagnosis of malignancy. There were no significant complications of sonographically guided core biopsy. CONCLUSION. Sonographically guided core biopsy is a highly accurate technique for evaluation of parotid lesions and can be safely performed as an outpatient procedure. Sonographically guided core biopsy has potential advantages over fine-needle aspiration cytologic examination, particularly in the typing and grading of lymphoma and carcinoma and in improved differentiation of reactive nodal hyperplasia from lymphoma. The use of sonographically guided core biopsy may help reduce the need for surgical biopsy and facilitates prompt referral to the appropriate clinical team. umerous pathologic processes present with parotid swelling, and it is N frequently difficult on clinical grounds alone to determine with confidence the nature of a parotid mass. Reaching an accurate diagnosis when contemplating surgery is important, because many nonneoplastic and some neoplastic lesions may not require surgical intervention, particularly in patients at poor anesthetic risk. It also may be possible to manage certain benign neoplasms, such as Warthin s tumor, with minimally invasive surgical techniques such as extracapsular dissection. Accurate diagnosis is essential for selection of the appropriate operative procedure and to allow informed patient consent. After the fall from favor of open biopsy as the result of a high level of tumor seeding [1], fine-needle aspiration cytology (FNAC) has been widely adopted for assessment of parotid masses. High diagnostic accuracy has been found in some studies [2 6]. FNAC has welldocumented limitations, however, and is frequently associated with high levels of inadequate or false-negative diagnoses [7, 8]. It is against this background that sonographically guided core biopsy of the parotid gland has been investigated [9 11]. Previous studies in- AJR:188, January

2 Howlett et al. Fig year-old woman with normal parotid gland. Longitudinal sonogram shows main intraparotid vessels. Retromandibular vein (large arrow) is superficial to external carotid artery (small arrow). Facial nerve is not shown but passes just superficial to retromandibular vein. Deep parotid lobe (D) is deep in relation to plane passing through path of nerve, and superficial lobe (S) is superficial to this plane. volved 16 [9], 54 [10], and 53 [11] patients with parotid swelling and showed a diagnostic accuracy of 97% [10, 11] to 100% [9] for sonographically guided core biopsy findings compared with surgical histologic findings, and there were no significant complications. In two of the studies [9, 10], 18-gauge or smaller needles were used with a mean of two passes per patient. In the third study [11], small- and largebore needles (up to 14-gauge) and a range of needle passes were used. The purpose of our study was to evaluate the accuracy and safety of sonographically guided core biopsy performed only with small-bore needles in a series of 135 patients presenting with a parotid mass. Subjects and Methods Over the 7-year period from 1998 to 2004, 135 patients consecutively referred for evaluation of a parotid mass were incorporated into this prospective study. Sixty-three patients were men, and 72 were women. The age range was years (mean, 60 years). After initial diagnostic sonographic examination (ATL HDI 5000 system, Philips Medical Systems) with a high-resolution linear array transducer (12 5 MHz; 7 4 MHz in larger patients or for deeper lesions), the operator proceeded to core biopsy. Before biopsy, written consent was obtained from the A patient, the benefits and potential hazards of the procedure having been explained. Contraindications to biopsy included inability to provide written consent or to cooperate and concurrent use of anticoagulant medication or presence of known bleeding diathesis. Preprocedural assessment of platelet count and clotting parameters was not undertaken routinely. All diagnostic sonographic examinations and biopsies were performed by a single operator using a freehand biopsy technique. With an aseptic technique and after administration of 1% lidocaine local anesthetic, a small skin incision was made and a biopsy needle introduced under sonographic guidance. Biopsy specimens were obtained with a spring-loaded device. For the first 16 patients a 22-mm fixed-throw mechanism (Biopty Gun, Bard) was used, and thereafter, a 15- to 22-mm variable throw device (Magnum gun, Bard). An average of two passes were made per patient. Biopsy samples were fixed in a jar containing 10% formalin and were sent for histologic analysis. All parotid lesions were demarcated sonographically as lying in the superficial lobe of the gland and had a size range of 6 58 mm. In most (n = 122) of the patients, 18-gauge needles were used. Because of the hazards of facial nerve and vascular injury, 20-gauge needles were used in patients (n = 13) with smaller lesions situated close to the main intraparotid vessels. Under sonographic guidance, lesions were approached along the longitudinal, oblique, or transverse plane, depending on the position of the lesion and to avoid adjacent structures (Fig. 1). Biopsy specimens of cystic lesions were obtained from the cyst wall, samples being directed at the more solid components of lesions of mixed cystic and solid appearance. The needle throw of the variable-throw biopsy device was selected according to the size and situation of the lesion to allow adequate lesion sampling but to ensure the needle did not exit deep in relation to the lesion (Fig. 2). Most of the biopsy procedures took less than 15 minutes. After biopsy, patients were asked to compress the puncture site and were observed for 30 minutes before discharge. Final diagnosis was established on the basis of adequate histologic findings in the core biopsy specimen and surgical histologic findings for the surgical group or clinical follow-up findings for patients who did not undergo surgery. The follow-up period ranged from a few months to 7 years. Results Of the 135 patients in the study, 76 underwent surgery and 59 avoided surgery. Initial sonograms showed a focal lesion suitable for biopsy in 124 patients; 11 patients in the nonsurgical group had diffuse sonographic abnormalities with no discrete focal lesion. All sonographically guided core biopsy specimens were considered satisfactory for evaluation by the histopathologist. In the surgical group, 55 patients had a diagnosis of benign neoplasm (37, pleomorphic adenoma [Fig. 3]; 16, Warthin s tumor [Fig. 4]; two, oncocytoma). Eighteen patients had a diagnosis of malignancy (three, mucoepidermoid carcinoma [Fig. 5]; one, adenoid cystic carci- Fig year-old woman with painless, progressive parotid gland enlargement and Sjögren s syndrome. Biopsy findings confirmed diagnosis of non-hodgkin s lymphoma complicating Sjögren s syndrome. A and B, Sonograms show 16-mm pseudocystic mass in tail of right parotid gland. Tip of biopsy needle (arrow, A) is positioned so that on needle discharge with 15-mm setting of biopsy device, needle traverses but does not exit lesion. Confirmation of needle placement is seen in B. B 224 AJR:188, January 2007

3 Sonographically Guided Biopsy of Parotid Mass Fig year-old man with painless parotid mass. Sonogram of pleomorphic adenoma shows rounded and circumscribed hypoechoic solid mass in superficial lobe of parotid. Distal acoustic enhancement is evident. Fig year-old woman with painless, rapidly enlarging parotid mass. Sonogram of mucoepidermoid carcinoma shows ill-defined, hypoechoic mass in superficial lobe of parotid gland. Lesion is of heterogeneous echotexture, and distal acoustic shadowing is present. Sonographic features are those of primary malignant tumor of parotid, although histologic subtypes cannot be differentiated with sonography. noma; two, adenocarcinoma; one, oncocytic carcinoma; two, squamous cell carcinoma; two, poorly differentiated carcinoma; one, melanoma; six, non-hodgkin s lymphoma). Three patients had miscellaneous, nonneoplastic diagnoses (one, lymphadenitis; one lymphoepithelial cyst; one, actinomycosis). In the nonsurgical group, 16 patients had a diagnosis of benign neoplasm (eight, pleomorphic adenoma; eight, Warthin s tumor), and 17 Fig year-old man with painless slow-growing parotid mass. Sonogram of Warthin s tumor shows circumscribed hypoechoic mass with inhomogeneous internal architecture containing solid and prominent cystic components. Fig year-old man with painless parotid gland enlargement. Longitudinal sonogram shows diffuse involvement of parotid gland in Sjögren s syndrome. Gland appears coarse and hypoechoic and contains multiple small hypoechoic foci that are in keeping with sialectatic changes of Sjögren s syndrome. had a diagnosis of malignant neoplasm (one, adenocarcinoma; one, small cell lung carcinoma; one, nasopharyngeal carcinoma; three, squamous cell carcinoma; 11, non-hodgkin s lymphoma). Fifteen patients had a nonneoplastic diagnosis and a focal lesion present at sonography (eight, reactive nodal hyperplasia; three, tuberculosis; one, sarcoidosis; one, retention cyst; two, sclerosing sialadenitis). Eleven patients had a nonneoplastic diagnosis and no focal lesion found at sonography (six, fatty infiltration; one, Sjögren s syndrome [Fig. 6]; one sarcoidosis, two chronic sialadenitis, one, normal parotid gland). The patient with a core biopsy diagnosis of normal parotid tissue presented with bilateral parotid swelling. At sonography the glands appeared hyperechoic and enlarged, suggestive of fatty infiltration. Biopsy specimens were obtained from the superficial lobe of the right parotid gland. After a normal histology report, no further investigations were undertaken. There was complete agreement between the results of sonographically guided core biopsy and surgical histologic findings in 74 of 76 patients. The two cases that did not correlate were malignant tumors. One of these tumors was diagnosed as squamous cell carcinoma at sonographically guided core biopsy, but the final histologic result was mucoepidermoid carcinoma. In the other case, the core biopsy diagnosis was mucoepidermoid carcinoma, but the final histologic result was squamous cell carcinoma. The treatment of these two patients was not affected by these findings; both underwent surgical excision. Only six of 17 patients with a core biopsy diagnosis of non-hodgkin s lymphoma needed surgical biopsy for further treatment information. The eight patients with reactive nodal hyperplasia diagnosed at sonographically guided core biopsy were well during clinical followup. Sixteen patients with benign neoplasms did not undergo surgical treatment. They either declined or were unfit for surgery and continued to undergo observation. Three nonneoplastic lesions were managed surgically either for symptom control or at patient request. Sonographically guided core biopsy was well tolerated by all patients with no immediate complications. A small subclinical hematoma was present at surgery in one patient who had undergone biopsy of a Warthin s tumor. The hematoma was in the soft tissue overlying the gland at the biopsy site and was impalpable. The patient had no predisposing risk factors for bleeding, and surgery was unaffected. No evidence of tumor seeding was detected during follow-up. To assess the utility of sonographically guided core biopsy in the diagnose of malignancy and to differentiate benign from malignant disease, we compared histologic results and sonographically guided core biopsy findings for the 76 patients who underwent both biopsy and surgery. There were 18 true-positive findings (malignant diagnosis) and 58 truenegative findings (benign diagnoses) and no AJR:188, January

4 Howlett et al. false-positive or false-negative findings, giving sonographically guided core biopsy a sensitivity, specificity, and diagnostic accuracy of 100% and 100% positive and negative predictive values in the diagnosis of malignancy. Discussion FNAC as a technique has a number of advantages in assessment of parotid masses. It is fast, safe, well tolerated, and accurate when performed by skilled practitioners. The accuracy of FNAC can be enhanced within a specialist clinic, which may be led by a cytologist [3] or a radiologist [12]. Outside the specialist clinic, however, the accuracy of FNAC decreases significantly with acquisition of large numbers of aspirate samples that do not give enough information for diagnosis [7, 8]. Even when the procedure is performed by an experienced practitioner, there are well-recognized pitfalls in the cytologic diagnosis of pleomorphic adenoma [13, 14] and Warthin s tumor [15]. Despite the use of ancillary cytologic techniques (flow cytometry and in situ hybridization) with FNAC in the diagnosis of lymphoid proliferation, it is often difficult to differentiate low-grade lymphoma from reactive nodal hyperplasia [16], and FNAC diagnosis of lymphoma is not generally considered definitive [17]. In these cases FNAC often acts as an indicator of the need for surgical biopsy. Differentiation of reactive atypia in benign squamous epithelium from well-differentiated squamous cell carcinoma may not be possible with FNAC [18]. In addition, it often is not possible to diagnose parotid involvement by systemic disease such as Sjögren s syndrome and sarcoidosis with FNAC or to accurately grade and type carcinomas and lymphomas or to differentiate in situ from invasive disease [10, 16]. FNAC has a low predictive value for benign nonneoplastic lesions [6], and a negative FNAC result has a low negative predictive value [5]. FNAC is more likely not to give enough information for diagnosis of small lesions, those of low cellularity, and those with uncommon histologic features. The pitfalls of FNAC have led some authors to study the use of sonographically guided core biopsy in assessment of parotid masses [9 11]. The combination of initial diagnostic sonography with biopsy is useful because the parotid glands are superficial structures, readily amenable to assessment with high-resolution sonography. Sonography is the initial imaging technique of choice for the characterization and delineation of parotid masses, and the findings can be used as a guide for additional imaging (usually MRI) of possibly malignant lesions and of lesions the full extent of which is difficult to assess, particularly if there is deep lobe involvement, an area of sonographic weakness due to mandibular obscuration [19, 20]. The use of sonographic guidance for biopsy is important because it allows accurate needle placement, such that the needle traverses but does not exit deep in relation to the lesion and avoids adjacent structures. A variable throw ability of the biopsy device is useful when space is confined. Some needle devices allow initial passage of the inner stylet for exact positioning of the needle tip, with subsequent discharge of the outer sheath over the fixed inner stylet to obtain a core of tissue. This technique may help reduce damage to deeper structures. Sonographic guidance also allows central sampling of a lesion with improved diagnostic yield [21], and necrotic areas can be bypassed. Sonography may also depict clinically impalpable disease. In our study, sonographically guided core biopsy had a diagnostic accuracy of 100% in the surgical group, comparable with previous findings. Sonographically guided core biopsy also had 100% sensitivity and specificity in the differentiation of benign from malignant disease and 100% positive and negative predictive values in the diagnosis of malignancy. There was no correlation between sonographically guided core biopsy findings and surgical histologic findings in only two cases, of mucoepidermoid and squamous cell carcinoma, but management was unaffected. In all cases, biopsy specimens with which a diagnosis could be determined were obtained with small-bore needles (18- or 20-gauge) with an average of two passes per patient. Sonographically guided core biopsy was well tolerated with no significant complications. Unlike FNAC, sonographically guided core biopsy requires formal histologic reporting, which is time-consuming, and it is more invasive than FNAC, requiring local anesthesia and a skin incision. The main advantage of core biopsy over FNAC is that a core of tissue is obtained that can be used for formal histologic and immunohistochemical analysis. This advantage improves differentiation of low-grade lymphoma from reactive nodal hyperplasia, allows grading and typing of lymphoma and poorly differentiated carcinoma, and helps determine the likely origin of metastatic carcinoma. In our study only six (35%) of 17 patients with non- Hodgkin s lymphoma needed surgical biopsy for additional treatment information, and no patient with a core biopsy diagnosis of reactive nodal hyperplasia underwent surgical excision, clinical follow-up sufficing. Eleven patients in whom no focal lesion was detected with sonography underwent sonographically guided core biopsy for evaluation of possible parotid involvement by systemic disease. Although there were no major complications of sonographically guided core biopsy in our study, there are hazards of core biopsy that need to be considered, including hemorrhage, facial nerve injury, and tumor seeding. The retromandibular vein and external carotid artery are the major intraparotid vessels; they are well visualized with sonography (Fig. 1) and can be avoided [20]. The facial nerve is not readily identified with sonography, although its position can be inferred because the nerve passes in a plane superficial to the retromandibular vein and can be avoided [14, 20]. Tumor seeding along the needle track is a hazard of both FNAC and core biopsy. The risk of tumor seeding varies according to the organ involved and the size of the needle used. Tumor seeding is described in as many as 12% of cases of FNAC of thoracic mesothelioma [22], with an incidence of only 0. 02% for abdominal tumors [23]. Evidence suggests that tumor seeding tends to occur with larger-bore needles, probably because these needles allow aspiration of sizable stromal fragments and allow survival of malignant cells [22]. Tumor seeding in the parotid gland is extremely rare after needle biopsy and has been described only in association with the use of large-bore needles [24]. Some surgeons excise the biopsy track, although this approach has not been adopted at our institution. Although there was no evidence of tumor seeding in our study population, patient follow-up was limited to only a few months in some cases, and continued surveillance is needed. In conclusion, sonographically guided core biopsy is a highly accurate, safe, and well-tolerated technique for the diagnosis of parotid mass. The procedure can be performed with small-bore needles on an outpatient basis. Sonographically guided core biopsy has advantages over FNAC, particularly in differentiation of lymphoma from reactive nodal hyperplasia, in typing and grading of carcinoma and lymphoma, and in the assessment of parotid involvement by systemic disease. The use of sonographically guided core biopsy may reduce the high rate of surgical biopsy associated with FNAC that does not provide enough information for diagnosis and may facilitate both prompt referral to the relevant 226 AJR:188, January 2007

5 Sonographically Guided Biopsy of Parotid Mass clinical team and selection of an appropriate operative procedure if needed. Acknowledgments We thank Carrie Favell for typing the manuscript, Nick Taylor for preparing the illustrations, and David Sallomi for assistance with the statistics. References 1. McGuirt WF, McCable BF. Significance of node biopsy before definitive treatment of cervical metastatic carcinoma. Laryngoscope 1978; 88: Eneroth CM, Franzen S, Zajicek J. Cytologic diagnosis on aspirate from 1000 salivary gland tumors. Acta Otolaryngol 1967; 1: Coghill S, Brown L Why pathologists should take needle aspiration specimens. Cytopathology 1991; 2: Filipoulos E, Angeli S, Daskalopoulou D, Kelessis N, Vassilopoulos P. Preoperative evaluation of parotid tumors by fine needle biopsy. Eur J Surg Onocol 1998; 24: Zbären P, Schär C, Hotz MA, Loosli H. Value of fine needle aspiration cytology of parotid gland masses. Laryngoscope 2001; 111: Cohen E G, Patel S G, Lin O, et al. Fine needle aspiration biopsy of salivary gland lesions in a selected patient population. Arch Otolaryngol Head Neck Surg 2004; 130: McGurk M, Hussain K, Role of fine needle aspiration cytology in the management of the discrete parotid lump. Ann R Coll Surg Engl 1997; 79: Balakrishnan K, Castling B, McMahan J, et al. Fine needle aspiration cytology in the management of parotid mass: a two centre retrospective study. Surgeon 2005; 2: Buckland JR, Manjaly G, Violaris N, Howlett DC. Ultrasound-guided cutting-needle biopsy of the parotid gland. J Laryngol Otol 1999; 113: Kesse KW, Manjaly G, Violaris N, Howlett DC. Ultrasound-guided biopsy in the evaluation of focal lesions and diffuse swelling of the parotid gland. Br J Oral Maxillofac Surg 2002; 40: Yung-Liang W, Siu-Cheung C, Yao-Liang C, et al. Sonography-guided core needle biopsy of parotid gland masses. Am J Neuroradiol 2004; 25: Robinson IA, Cozens NJ. Does a joint ultrasoundguided cytology clinic optimize the cytologic evaluation of head and neck masses? Clin Radiol 1999; 54: Brachtel EF, Pilch BZ, Khettry V, et al. Fine needle aspiration biopsy of a cystic pleomorphic adenoma with extensive adnexa-like differentiation: differential diagnostic pitfall with mucoepidermoid carcinoma. Diagn Cytopathol 2003; 28: Verma K, Kapila K. Role of fine needle aspiration cytology in the diagnosis of pleomorphic adenoma. Cytopathology 2002; 13: Parwarni AV, Ali-Sayed Z. Diagnostic accuracy and pitfalls in the fine needle aspiration interpretation of Warthin s tumor. Cancer 2003; 99: Loggins JP, Urquhart A. Preoperative distinction of parotid lymphomas. J Am Coll Surg 2004; 199: Lioe TF, Elliot H, Allen DC, Spence RA. The role of the fine needle aspiration cytology (FNAC) in the investigation of superficial lymphadenopathy: uses and limitations of the technique. Cytopathology 1999; 10: Layfield L. Fine needle aspiration of the head and neck. Pathology 1996; 4: Howlett DC, Kesse KW, Hughes DV, Sallomi DF. The role of imaging in the evaluation of parotid disease. Clin Radiol 2002; 57: Howlett DC. High resolution ultrasound assessment of the parotid gland. Br J Radiol 2003; 76: McIvor NP, Freeman JL, Salem S, et al. Ultrasound and ultrasound-guided fine-needle aspiration biopsy of head and neck lesions: a surgical perspective. Laryngoscope 1994; 104: Sinner WN, Zajicek J. Implantation metastasis after percutaneous transthoracic needle aspiration biopsy. Acta Radiol Diagn 1980; 17: Roussel F, Dalion J, Benozio M. The risk of tumor seeding in needle biopsies. Acta Cytol 1989; 33: Yamaguchi KT, Strong MS, Shapshay SM, Soto E. Seeding of parotid carcinoma along Vim-Silverman needle tract. J Otolaryngol 1979; 8:49 52 AJR:188, January