FINE=NEEDLE ASPIRATION OF PAROTID GLAND LESIONS Mark S. Weinberger, MD, Wouter W. Rosenberg, William T. Meurer, CT (ASCP), and K. Thomas Robbins, MD
The results of 49 specimens obtained by fine-needle aspiration biopsy of parotid gland lesions were compared with the pathologic diagnoses of the surgically resected specimens. Cytologically, 33 lesions were diagnosed as benign, with 30 of these confirmed histologically and three false-negative results. Fourteen cytologic specimens were called malignant or suspicious for malignancy, with 11 of these confirmed histologically and three false-positiveresults. The concurrence rate for distinguishing benign from malignant disease was 87.2%. The sensitivity for malignancy was 78.6% and the specificity 90.9%. The pathology of the misdiagnosed lesions will be reviewed. Based on our data and a review of the literature, we conclude that fine needle aspiration biopsy of parotid gland masses, with the observation of certain caveats, is a helpful adjunctive test for diagnosis and treatment planning. HEAD & NECK 1992;14:483-487 0 1992 John Wiley & Sons, Inc.
Aspiration cytology as a method for tumor diagnosis was first reported by Kun' in 1847. This technique was first advocated for the diagnosis of
From the Division of Otolaryngology-Head and Neck Surgery (Drs. Weinberger and Robbins and Mr. Rosenberg), and Department of Pathology (Mr. Meurer). UCSD Medical Center, San Diego, California. Presented at the 75th annual meeting of The Pacific Coast Oto-Ophthalmologic Society, Monterey, California, June 15-19, 1991. Address reprint requests to Dr. Robbins at Director, Head and Neck Oncology, UCSD Cancer Center, 220 Dickinson Street, San Diego, CA 92103. Accepted for publication May 13, 1992. CCC 0148-6403/92/060483-05 0 1992 John Wiley & Sons, Inc.
Parotid FNA
head and neck tumors by Martin and Ellis2 in 1930, but rapidly fell into disfavor because of complications, such as bleeding and tumor seeding, resulting from the use of large-bore needles. Since the advent of exfoliative cytology and the use of fine needles, many studies have shown it to be a safe, accurate, and cost-effective method for evaluating neck m a ~ s e s . ~In- ~ contrast, the use of fine-needle aspiration (FNA) cytology in the diagnosis of parotid masses has been fraught with controversy. The reasons for the disparity of opinion have included fear of facial nerve injury or tumor seeding, as well as a lack of uniform criteria for diagnosis and lack of confidence in the diagnostic accuracy of the cytologic smear^.^,^ In the 1950s and 1960s, pioneering work by Eneroth and colleagues at the Karolinska Institute in Stockholm laid the groundwork for the establishment of uniform cytologic criteria for the diagnosis of salivary tumors and for the acceptance of this technique as a safe and cost-effective method for preoperative d i a g n ~ s i s . ~ * ' - ~ ~ Diversity of opinion regarding the impact of parotid FNA on clinical decision-making has persisted, mostly owing to the high false-negative rate and thus lower sensitivity in comparison to head and neck tumors in general.14 We reviewed our experience for a 5-year period with FNA of parotid masses for the specific purpose of determining its accuracy and thus helping to better define the role of this technique in patient management.
HEAD & NECK
NovembedDecernber 1992
483
TP TP+FN
MATERIALS AND METHODS
Between 1985 and 1989, 49 patients underwent fine needle aspiration biopsy (FNAB) of a parotid mass with subsequent surgical resection at the University of California Medical Center, San Diego, and the San Diego VA Medical Center. The study population consisted of 40 men and nine women with a mean age of 52 years. All patients presented with a mass in the parotid region. Only patients who underwent both FNAB and parotidectomy were included in this retrospective review. Our technique of FNA is essentially the same as that described by 2 a j i ~ e k . The l ~ skin is prepped with an alcohol wipe and the mass is grasped between the thumb and forefinger of the left hand. A 22-gauge needle attached to a 20-mL syringe in a syringe pistol is inserted into the mass and suction is applied to the aspirating syringe. The needle is then moved back and forth in different directions or passed multiple times through the same tract. The pressure is released and then the needle is withdrawn. The aspirate is smeared on glass slides. One slide is air-dried for Wright’s stain, and the other is immersed in 95% ethanol for Papanicolaou stain. The needle, is then rinsed in Hanks’ solution, which is spun down and filtered, and then stained. In all cases, the cytologic diagnoses based on the FNA smears were reviewed and compared to the final pathologic diagnoses based on permanent histologic sections. For the purposes of this study the FNA results were classified into several categories: True-positive: The presence of a malignancy was correctly diagnosed cytologically (this includes smears which were read as “suspicious for malignancy”). True-negative: The absence of a malignancy was correctly diagnosed cytologically. False-positive: The cytologic smear was considered diagnostic or suspicious for malignancy in the presence of a benign process. False-negative: The cytologic smear failed to diagnose a malignancy. Nondiagnostic: A cytologic diagnosis could not be made on the basis of the material obtained. The sensitivity for malignancy was calculated as:
484
Parotid FNA
The specificity was calculated as:
FN TN+FP The accuracy of differentiation (benign from malignant) as:
TP+TN Total RESULTS
Forty-nine FNA specimens were obtained for cytologic evaluation. Two of these contained insufficient cellular material for diagnosis. Thirtythree specimens were reported to be benign based on cytology. Subsequently, on histologic analysis of the final pathologic specimens, 30 of these 33 were confirmed to be benign lesions (Table 1). The three false-negative FNAs were a papillary cystic variant of acinic carcinoma diagnosed cytologically as Warthin’s tumor (Table 2, Figures lA, lB, and a), an acinic carcinoma diagnosed as sialadenitis, and a neurofibrosarcoma diagnosed as a pleomorphic adenoma. Fourteen FNA specimens were reported to be malignant or suspicious for malignancy. Eleven of these were confirmed malignancies on final pathology (Table 3). The three false-positive results consisted of two pleomorphic adenomas, diagnosed on cytologic smears as “possible malignant transformation” and “mucoepidermoid carcinoma favored,” and one case of inflammation and lymphoid hyperplasia diagnosed cytologically as “consistent with lymphoma” (Table 2). In our series, the accuracy of differentiation of benign versus malignant tumors was 87.2%. The sensitivity for malignancy was 78.6% and the specificity 90.9%. The false-negative and false-positive rates were both 6.1%. DISCUSSION
Our results are consistent with previously published series of salivary gland FNA. The concurrence rate for differentiating benign from malignant lesions has varied from 81%16 to 98%,17 with our concurrence rate of 87.2% falling well within that range. Our false-positive rate of 6.1%
HEAD & NECK
NovembedDecernber 1992
Table 1. True negative FNA diagnosis for the absence of malignancy. Patient (age, yrlsex) 161M 53li 58lF 37lM 13lM 25lM 24lM 301M 56lM 58lF 30lF 39lM 49lF 44lF 35lF 69lM 40lM 49lM 33lM 65lM 63lM 691M 58lM 56lM 66lM 62lM 531M
4a1~ 771M 63lM
FNA diagnosis
Pathologic diagnosis
Pleomorphic adenoma No atypical cells Pleomorphic adenoma Pleomorphic adenoma Reactive lymph node Sialadenitis Pleomorphic adenoma Benign cyst Warthin's tumor Benign cyst Pleornorphic adenoma Pleomorphic adenoma No atypical cells Warthin's tumor Pleornorphic adenoma Possible warthin's tumor Pleomorphic adenoma No atypical cells Normal lymphocytes Pleomorphic adenoma No atypical cells Negative for malignancy Negative for malignancy Histiocytes, no atypia Negative for malignancy Warthin's tumor Negative for malignancy Possible oncocytoma Negative for malignancy Negative for malignancy
Pleomorphic adenoma Sialadenitis Pleomorphic adenoma Monomorphic adenoma Chronic inflammation Chronic inflammation Pleomorphic adenoma Branchial cyst Warth i n's tu mor Warthin's tumor Pleomorphic adenoma Pleomorphic adenoma Chronic inflammation Warthin's tumor Pleomorphic adenoma Oncocytoma Pleomorphic adenoma Benign cyst Hyperplastic lymph node Pleomorphic adenoma Lipoma Pleomorphic adenorna Monomorphic adenoma Lymphoepithelial cyst Normal tissue Warthin's tumor Lyrnphadenitis Warthin's tumor Warthin's tumor Warthin's tumor
was slightly higher than previous reports, which range from 0%l7 to 4 . ~ 0 . ~ ~ Two of our three false-positive diagnoses resulted from the overdiagnosis of cellular pleomorphic adenomas, one as a mucoepidermoid carcinoma and the other as a possible malignant transformation of a pleomorphic adenoma. Overdiagnosis of these highly cellular lesions with mucoid elements and a paucity of stroma is the
most common cause of false-positive results in salivary gland FNA.14,18Layfield et a1.18 suggest that awareness of this common error can minimize its occurrence, and they propose stricter criteria of atypical mucinous and squamous components, nuclear irregularity, and prominent nucleoli to justify a diagnosis of mucoepidermoid carcinoma. Our other false-positive result was a case of
Tsbie 2. False-positive and false-negative FNA diagnoses. Patient (age, yrlsex) 42lM 581F 55lM 35lM 121i 201M
Parotid FNA
FNA diagnosis
Pathologic diagnosis
Pleomorphic adenoma, possible malignant transformation Mucoepidermoid carcinoma Consistent with lymphoma Warthin's tumor Sialadenitis Pleomorphic adenoma
Cellular pleomorphic adenoma Cellular pleornorphic adenoma Lymphoid hyperplasia Acinic carcinoma (papillary cystic variant) Acinic cell carcinoma Neurofibrosarcoma
HEAD & NECK
NovemberlDecernber 1992
485
A
B FIGURE 1. Histologic section of papillary cystic variant of acinic cell carcinoma. (A) Low-power view shows papillary and cystic structures (hematoxylin & eosin, original magnification x40). (B) High-power view shows typical malignant acinar cells with granular and clear cytoplasm, significant atypia, and mitotic figures (hernatoxylin & eosin, original magnification x200).
inflammation and lymphoid hyperplasia, which was diagnosed by FNA as “consistent with lymphoma.” This contrasts with previous reports in which monomorphous lymphoid infiltrates have been responsible for false-negative results with the underdiagnosis of lymphomas.14~18~19Layfield et a1.I8 suggest the use of immunohistochemical markers to determine the monoclonality of these lesions. Much of the controversy concerning the clinical value of salivary gland FNA has arisen from historically high false-negative rates. As reviewed by Layfield et al., these rates have ranged from as high as 37%6 in early studies to
FIGURE 2. High-power view of FNA smear shows papillary groupings of cells which were incorrectly interpreted as typical oncocytic cells of a Warthin’s tumor (Papanicalaou, original magnification x400).
486
Parotid FNA
an average of less than 10% in more recent series. This is mostly due to the pioneering work of Zajicek and Eneroth in developing uniform criteria for the diagnosis of salivary gland cytologic smears.8 Our false-negative rate of 6.1% compares favorably with most published series. Two of our three false-negative results were acinic cell carcinomas, which were misdiagnosed as Warthin’s tumor and sialadenitis. In other published s e r i e ~ , ’ ~acinic ’ ~ ~ cell carcinoma has been infrequently misdiagnosed since Zajicek and Eneroth established the diagnostic criteria of the presence of numerous acinic cells in solid plugs with total absence of ductal epithelium. However, our first false-negative was a case of a papillary cystic variant of acinic cell carcinoma, and the cytologic diagnosis was based on papillary groupings and sheets of oncocytic cells in a background of cystic degeneration, features most consistent with Warthin’s tumor. The other misdiagnosed acinic cell carcinoma is a well-described form of sampling error which is a common cause of false-negative results. Small tumors can cause ductal obstruction and secondary ~ialadenitis.~ The third false-negative result was a neurofibrosarcoma which was misdiagnosed as a pleomorphic adenoma based on a mixed population of stromal and epithelial components in the cytology smear. This most likely represents a sampling error. FNA cytology of parotid gland masses is not as accurate as FNA of thyroid tumors or squamous cell carcinoma of the head and neck. However, with a diagnostic accuracy for differentiating be-
HEAD & NECK
Novernber/December 1992
Table 3. True-positive FNA diagnoses for the presence of malignancy. Patient (age, yrlsex)
FNA diagnosis
Pathologic diagnosis Squamous cell carcinoma Squamous cell carcinoma Lymphoma
46lF 82lM 55lM
Squamous cell carcinoma Squamous cell carcinoma Atypical lymphocytes, suspicious for malignancy Adenocarcinoma Carcinoma suspected Squamous cell carcinoma
66lM
Malignant cells
61lM
Squarnous cell carcinoma
75lM 61lM
Malignancy not otherwise specified Atypical epithelial cells
62lM
Merkel cell carcinoma
51lM 59lM
501M
4. 5. 6. 7.
Adenocarcinoma Melanoma Squamous cell carcinoma Undifferentiated malignancy Squarnous cell carcinoma Adenocarcr noma Squamous cell carcinoma Merkel cell carcinoma
nign from malignant lesions, which is consistently greater than 85% in previously published studies and the present stud , it is ver useful in certain clinical situations. JNA may e used t o localize tumors to the parotid gland, as it is very accurate in differentiating metastatic squamous cell carcinoma from a primary parotid tumor. It is also very useful in following parotid masses in patients who are poor o erative risks. Finally, the routine use of paroti FNA will contribute t o improve preoperative planning. As this review of our experience with parotid FNA has shown, the head and neck surgeon who performs parotid FNA must be aware of the histologic criteria for the various tumors, and the limitations on the diagnosis of each one. Excellent communication between surgeon and cytopathologist, and between surgeon and patient is essential for the proper use of this test. When these principles are adhered to, FNA contributes greatly to the management of parotid tumors.
i
2
REFERENCES 1. Kun M. A new instrument for the diagnosis of tumors. Month J Med Sci 1847;7:853-854.
Parotid FNA
2. Martin H, Ellis EB. Biopsy of needle puncture and aspiration. Ann Surg 1930;92:169-181. 3. Schwarz R, Chan NH, MacFarlane JK. Fine needle aspi-
ration cytology in the evaluation of head and neck masses. A m J Surg 1990;159:482-485. Frable MAS, Frable WJ. Fine needle aspiration biopsy revisited. Laryngoscope 1982;92:1414- 1418. Engzell U, Jakobsson PA, Siguardson A, Zajicek J. Aspiration biopsy of metastatic carcinoma in lymph nodes of the neck. Acta Otolaryngol 1971;72:138- 147. Eneroth CM, Franzen S, Zajicek J . Cytologic diagnosis on aspirate from 1000 salivary gland tumors. Acta Otolaryngol [Suppl] 1966;224:168- 171. Rodriguez HP, Silver CE, Moisa 11, Chacho MS. Fineneedle aspiration of parotid tumors. A m J Surgy
1989;158:342-344. 8. Zajicek J, Eneroth CM. Cytological diagnosis of salivary
gland carcinomata from aspiration biopsy smears. Actu Otolaryngol 1970;263:183- 185. 9. Mavec P, Eneroth CM, Franzen S, Moberger G, Zajicek J. Aspiration biopsy of salivary gland tumors. Acta Otolaryngol 1964;58:417-484. 10. Eneroth Cm, Zajicek J. Aspiration biopsy of salivary gland tumors 111: morphologic studies on smears and histologic sections from 368 mixed tumors. Acta Cytol 1966;10:440-453. 11. Eneroth CM, Zajicek J . Aspiration biopsy of salivary
gland tumors IV: morphologic studies on smears and histologic sections from 45 cases of adenoid cystic carcinoma. Acta Cytol 1969;13:59-63. 12. Eneroth CM, Zajicek J. Aspiration biopsy of salivary gland tumors V: morphologic studies on smears and histologic sections of acinic cell carcinoma. Actu Rudiol [Suppl] 1971;310:85 - 93. 13. Eneroth CM, Zajicek J. Aspiration biopsy of salivary gland tumors VI: morphologic studies on smears and histologic sections from mucoepidermoid carcinoma. Acta Cytol 1966;20:35-41. 14. O'Dwyer P, Farrar WB, James AG, Finkelmeier W, McCabe DP. Needle aspiration biopsy of major salivary gland tumors: its value. Cancer 1986;57:554557. 15. Zajicek J. Aspiration biopsy cytology, part I: cytol-
ogy of supra-diaphragmatic organs. In: (series) Monographs in clinical cytology, vol. 4. New York: S. Karger, 1974~1-5. 16. Lindberg LG, Akeramn M. Aspiration cytology of sali-
vary gland tumors: diagnostic experience from six years of routine laboratory work. Laryngoscope 1976;86:584594. 17. Qizilbash AH, Sianol J, Young JE, et al. Fine needle as-
piration biopsy cytology of major salivary glands. Actu Cytol 1985;29:503-512. 18. Layfield LJ, Tan P, Glasgow BJ. Fine-needle aspiration of salivary gland lesions. Arch Pathol Lab Med 1987; 111:346-353. 19. Bono A, Chiesa F, Sala L, et al. Fine-needle aspiration biopsy in parotid masses. Tumori 1983;69:417421. 20. Persson PS, Zettergren L. Cytologic diagnosis of salivary gland tumors by aspiration biopsy. Actu Cytol 1973; 17~351-354.
HEAD & NECK
NovemberlDecember 1992
487
The results of 49 specimens obtained by fine-needle aspiration biopsy of parotid gland lesions were compared with the pathologic diagnoses of the surgically resected specimens. Cytologically, 33 lesions were diagnosed as benign, with 30 of these confirmed histologically and three false-negative results. Fourteen cytologic specimens were called malignant or suspicious for malignancy, with 11 of these confirmed histologically and three false-positiveresults. The concurrence rate for distinguishing benign from malignant disease was 87.2%. The sensitivity for malignancy was 78.6% and the specificity 90.9%. The pathology of the misdiagnosed lesions will be reviewed. Based on our data and a review of the literature, we conclude that fine needle aspiration biopsy of parotid gland masses, with the observation of certain caveats, is a helpful adjunctive test for diagnosis and treatment planning. HEAD & NECK 1992;14:483-487 0 1992 John Wiley & Sons, Inc.
Aspiration cytology as a method for tumor diagnosis was first reported by Kun' in 1847. This technique was first advocated for the diagnosis of
From the Division of Otolaryngology-Head and Neck Surgery (Drs. Weinberger and Robbins and Mr. Rosenberg), and Department of Pathology (Mr. Meurer). UCSD Medical Center, San Diego, California. Presented at the 75th annual meeting of The Pacific Coast Oto-Ophthalmologic Society, Monterey, California, June 15-19, 1991. Address reprint requests to Dr. Robbins at Director, Head and Neck Oncology, UCSD Cancer Center, 220 Dickinson Street, San Diego, CA 92103. Accepted for publication May 13, 1992. CCC 0148-6403/92/060483-05 0 1992 John Wiley & Sons, Inc.
Parotid FNA
head and neck tumors by Martin and Ellis2 in 1930, but rapidly fell into disfavor because of complications, such as bleeding and tumor seeding, resulting from the use of large-bore needles. Since the advent of exfoliative cytology and the use of fine needles, many studies have shown it to be a safe, accurate, and cost-effective method for evaluating neck m a ~ s e s . ~In- ~ contrast, the use of fine-needle aspiration (FNA) cytology in the diagnosis of parotid masses has been fraught with controversy. The reasons for the disparity of opinion have included fear of facial nerve injury or tumor seeding, as well as a lack of uniform criteria for diagnosis and lack of confidence in the diagnostic accuracy of the cytologic smear^.^,^ In the 1950s and 1960s, pioneering work by Eneroth and colleagues at the Karolinska Institute in Stockholm laid the groundwork for the establishment of uniform cytologic criteria for the diagnosis of salivary tumors and for the acceptance of this technique as a safe and cost-effective method for preoperative d i a g n ~ s i s . ~ * ' - ~ ~ Diversity of opinion regarding the impact of parotid FNA on clinical decision-making has persisted, mostly owing to the high false-negative rate and thus lower sensitivity in comparison to head and neck tumors in general.14 We reviewed our experience for a 5-year period with FNA of parotid masses for the specific purpose of determining its accuracy and thus helping to better define the role of this technique in patient management.
HEAD & NECK
NovembedDecernber 1992
483
TP TP+FN
MATERIALS AND METHODS
Between 1985 and 1989, 49 patients underwent fine needle aspiration biopsy (FNAB) of a parotid mass with subsequent surgical resection at the University of California Medical Center, San Diego, and the San Diego VA Medical Center. The study population consisted of 40 men and nine women with a mean age of 52 years. All patients presented with a mass in the parotid region. Only patients who underwent both FNAB and parotidectomy were included in this retrospective review. Our technique of FNA is essentially the same as that described by 2 a j i ~ e k . The l ~ skin is prepped with an alcohol wipe and the mass is grasped between the thumb and forefinger of the left hand. A 22-gauge needle attached to a 20-mL syringe in a syringe pistol is inserted into the mass and suction is applied to the aspirating syringe. The needle is then moved back and forth in different directions or passed multiple times through the same tract. The pressure is released and then the needle is withdrawn. The aspirate is smeared on glass slides. One slide is air-dried for Wright’s stain, and the other is immersed in 95% ethanol for Papanicolaou stain. The needle, is then rinsed in Hanks’ solution, which is spun down and filtered, and then stained. In all cases, the cytologic diagnoses based on the FNA smears were reviewed and compared to the final pathologic diagnoses based on permanent histologic sections. For the purposes of this study the FNA results were classified into several categories: True-positive: The presence of a malignancy was correctly diagnosed cytologically (this includes smears which were read as “suspicious for malignancy”). True-negative: The absence of a malignancy was correctly diagnosed cytologically. False-positive: The cytologic smear was considered diagnostic or suspicious for malignancy in the presence of a benign process. False-negative: The cytologic smear failed to diagnose a malignancy. Nondiagnostic: A cytologic diagnosis could not be made on the basis of the material obtained. The sensitivity for malignancy was calculated as:
484
Parotid FNA
The specificity was calculated as:
FN TN+FP The accuracy of differentiation (benign from malignant) as:
TP+TN Total RESULTS
Forty-nine FNA specimens were obtained for cytologic evaluation. Two of these contained insufficient cellular material for diagnosis. Thirtythree specimens were reported to be benign based on cytology. Subsequently, on histologic analysis of the final pathologic specimens, 30 of these 33 were confirmed to be benign lesions (Table 1). The three false-negative FNAs were a papillary cystic variant of acinic carcinoma diagnosed cytologically as Warthin’s tumor (Table 2, Figures lA, lB, and a), an acinic carcinoma diagnosed as sialadenitis, and a neurofibrosarcoma diagnosed as a pleomorphic adenoma. Fourteen FNA specimens were reported to be malignant or suspicious for malignancy. Eleven of these were confirmed malignancies on final pathology (Table 3). The three false-positive results consisted of two pleomorphic adenomas, diagnosed on cytologic smears as “possible malignant transformation” and “mucoepidermoid carcinoma favored,” and one case of inflammation and lymphoid hyperplasia diagnosed cytologically as “consistent with lymphoma” (Table 2). In our series, the accuracy of differentiation of benign versus malignant tumors was 87.2%. The sensitivity for malignancy was 78.6% and the specificity 90.9%. The false-negative and false-positive rates were both 6.1%. DISCUSSION
Our results are consistent with previously published series of salivary gland FNA. The concurrence rate for differentiating benign from malignant lesions has varied from 81%16 to 98%,17 with our concurrence rate of 87.2% falling well within that range. Our false-positive rate of 6.1%
HEAD & NECK
NovembedDecernber 1992
Table 1. True negative FNA diagnosis for the absence of malignancy. Patient (age, yrlsex) 161M 53li 58lF 37lM 13lM 25lM 24lM 301M 56lM 58lF 30lF 39lM 49lF 44lF 35lF 69lM 40lM 49lM 33lM 65lM 63lM 691M 58lM 56lM 66lM 62lM 531M
4a1~ 771M 63lM
FNA diagnosis
Pathologic diagnosis
Pleomorphic adenoma No atypical cells Pleomorphic adenoma Pleomorphic adenoma Reactive lymph node Sialadenitis Pleomorphic adenoma Benign cyst Warthin's tumor Benign cyst Pleornorphic adenoma Pleomorphic adenoma No atypical cells Warthin's tumor Pleornorphic adenoma Possible warthin's tumor Pleomorphic adenoma No atypical cells Normal lymphocytes Pleomorphic adenoma No atypical cells Negative for malignancy Negative for malignancy Histiocytes, no atypia Negative for malignancy Warthin's tumor Negative for malignancy Possible oncocytoma Negative for malignancy Negative for malignancy
Pleomorphic adenoma Sialadenitis Pleomorphic adenoma Monomorphic adenoma Chronic inflammation Chronic inflammation Pleomorphic adenoma Branchial cyst Warth i n's tu mor Warthin's tumor Pleomorphic adenoma Pleomorphic adenoma Chronic inflammation Warthin's tumor Pleomorphic adenoma Oncocytoma Pleomorphic adenoma Benign cyst Hyperplastic lymph node Pleomorphic adenoma Lipoma Pleomorphic adenorna Monomorphic adenoma Lymphoepithelial cyst Normal tissue Warthin's tumor Lyrnphadenitis Warthin's tumor Warthin's tumor Warthin's tumor
was slightly higher than previous reports, which range from 0%l7 to 4 . ~ 0 . ~ ~ Two of our three false-positive diagnoses resulted from the overdiagnosis of cellular pleomorphic adenomas, one as a mucoepidermoid carcinoma and the other as a possible malignant transformation of a pleomorphic adenoma. Overdiagnosis of these highly cellular lesions with mucoid elements and a paucity of stroma is the
most common cause of false-positive results in salivary gland FNA.14,18Layfield et a1.18 suggest that awareness of this common error can minimize its occurrence, and they propose stricter criteria of atypical mucinous and squamous components, nuclear irregularity, and prominent nucleoli to justify a diagnosis of mucoepidermoid carcinoma. Our other false-positive result was a case of
Tsbie 2. False-positive and false-negative FNA diagnoses. Patient (age, yrlsex) 42lM 581F 55lM 35lM 121i 201M
Parotid FNA
FNA diagnosis
Pathologic diagnosis
Pleomorphic adenoma, possible malignant transformation Mucoepidermoid carcinoma Consistent with lymphoma Warthin's tumor Sialadenitis Pleomorphic adenoma
Cellular pleomorphic adenoma Cellular pleornorphic adenoma Lymphoid hyperplasia Acinic carcinoma (papillary cystic variant) Acinic cell carcinoma Neurofibrosarcoma
HEAD & NECK
NovemberlDecernber 1992
485
A
B FIGURE 1. Histologic section of papillary cystic variant of acinic cell carcinoma. (A) Low-power view shows papillary and cystic structures (hematoxylin & eosin, original magnification x40). (B) High-power view shows typical malignant acinar cells with granular and clear cytoplasm, significant atypia, and mitotic figures (hernatoxylin & eosin, original magnification x200).
inflammation and lymphoid hyperplasia, which was diagnosed by FNA as “consistent with lymphoma.” This contrasts with previous reports in which monomorphous lymphoid infiltrates have been responsible for false-negative results with the underdiagnosis of lymphomas.14~18~19Layfield et a1.I8 suggest the use of immunohistochemical markers to determine the monoclonality of these lesions. Much of the controversy concerning the clinical value of salivary gland FNA has arisen from historically high false-negative rates. As reviewed by Layfield et al., these rates have ranged from as high as 37%6 in early studies to
FIGURE 2. High-power view of FNA smear shows papillary groupings of cells which were incorrectly interpreted as typical oncocytic cells of a Warthin’s tumor (Papanicalaou, original magnification x400).
486
Parotid FNA
an average of less than 10% in more recent series. This is mostly due to the pioneering work of Zajicek and Eneroth in developing uniform criteria for the diagnosis of salivary gland cytologic smears.8 Our false-negative rate of 6.1% compares favorably with most published series. Two of our three false-negative results were acinic cell carcinomas, which were misdiagnosed as Warthin’s tumor and sialadenitis. In other published s e r i e ~ , ’ ~acinic ’ ~ ~ cell carcinoma has been infrequently misdiagnosed since Zajicek and Eneroth established the diagnostic criteria of the presence of numerous acinic cells in solid plugs with total absence of ductal epithelium. However, our first false-negative was a case of a papillary cystic variant of acinic cell carcinoma, and the cytologic diagnosis was based on papillary groupings and sheets of oncocytic cells in a background of cystic degeneration, features most consistent with Warthin’s tumor. The other misdiagnosed acinic cell carcinoma is a well-described form of sampling error which is a common cause of false-negative results. Small tumors can cause ductal obstruction and secondary ~ialadenitis.~ The third false-negative result was a neurofibrosarcoma which was misdiagnosed as a pleomorphic adenoma based on a mixed population of stromal and epithelial components in the cytology smear. This most likely represents a sampling error. FNA cytology of parotid gland masses is not as accurate as FNA of thyroid tumors or squamous cell carcinoma of the head and neck. However, with a diagnostic accuracy for differentiating be-
HEAD & NECK
Novernber/December 1992
Table 3. True-positive FNA diagnoses for the presence of malignancy. Patient (age, yrlsex)
FNA diagnosis
Pathologic diagnosis Squamous cell carcinoma Squamous cell carcinoma Lymphoma
46lF 82lM 55lM
Squamous cell carcinoma Squamous cell carcinoma Atypical lymphocytes, suspicious for malignancy Adenocarcinoma Carcinoma suspected Squamous cell carcinoma
66lM
Malignant cells
61lM
Squarnous cell carcinoma
75lM 61lM
Malignancy not otherwise specified Atypical epithelial cells
62lM
Merkel cell carcinoma
51lM 59lM
501M
4. 5. 6. 7.
Adenocarcinoma Melanoma Squamous cell carcinoma Undifferentiated malignancy Squarnous cell carcinoma Adenocarcr noma Squamous cell carcinoma Merkel cell carcinoma
nign from malignant lesions, which is consistently greater than 85% in previously published studies and the present stud , it is ver useful in certain clinical situations. JNA may e used t o localize tumors to the parotid gland, as it is very accurate in differentiating metastatic squamous cell carcinoma from a primary parotid tumor. It is also very useful in following parotid masses in patients who are poor o erative risks. Finally, the routine use of paroti FNA will contribute t o improve preoperative planning. As this review of our experience with parotid FNA has shown, the head and neck surgeon who performs parotid FNA must be aware of the histologic criteria for the various tumors, and the limitations on the diagnosis of each one. Excellent communication between surgeon and cytopathologist, and between surgeon and patient is essential for the proper use of this test. When these principles are adhered to, FNA contributes greatly to the management of parotid tumors.
i
2
REFERENCES 1. Kun M. A new instrument for the diagnosis of tumors. Month J Med Sci 1847;7:853-854.
Parotid FNA
2. Martin H, Ellis EB. Biopsy of needle puncture and aspiration. Ann Surg 1930;92:169-181. 3. Schwarz R, Chan NH, MacFarlane JK. Fine needle aspi-
ration cytology in the evaluation of head and neck masses. A m J Surg 1990;159:482-485. Frable MAS, Frable WJ. Fine needle aspiration biopsy revisited. Laryngoscope 1982;92:1414- 1418. Engzell U, Jakobsson PA, Siguardson A, Zajicek J. Aspiration biopsy of metastatic carcinoma in lymph nodes of the neck. Acta Otolaryngol 1971;72:138- 147. Eneroth CM, Franzen S, Zajicek J . Cytologic diagnosis on aspirate from 1000 salivary gland tumors. Acta Otolaryngol [Suppl] 1966;224:168- 171. Rodriguez HP, Silver CE, Moisa 11, Chacho MS. Fineneedle aspiration of parotid tumors. A m J Surgy
1989;158:342-344. 8. Zajicek J, Eneroth CM. Cytological diagnosis of salivary
gland carcinomata from aspiration biopsy smears. Actu Otolaryngol 1970;263:183- 185. 9. Mavec P, Eneroth CM, Franzen S, Moberger G, Zajicek J. Aspiration biopsy of salivary gland tumors. Acta Otolaryngol 1964;58:417-484. 10. Eneroth Cm, Zajicek J. Aspiration biopsy of salivary gland tumors 111: morphologic studies on smears and histologic sections from 368 mixed tumors. Acta Cytol 1966;10:440-453. 11. Eneroth CM, Zajicek J . Aspiration biopsy of salivary
gland tumors IV: morphologic studies on smears and histologic sections from 45 cases of adenoid cystic carcinoma. Acta Cytol 1969;13:59-63. 12. Eneroth CM, Zajicek J. Aspiration biopsy of salivary gland tumors V: morphologic studies on smears and histologic sections of acinic cell carcinoma. Actu Rudiol [Suppl] 1971;310:85 - 93. 13. Eneroth CM, Zajicek J. Aspiration biopsy of salivary gland tumors VI: morphologic studies on smears and histologic sections from mucoepidermoid carcinoma. Acta Cytol 1966;20:35-41. 14. O'Dwyer P, Farrar WB, James AG, Finkelmeier W, McCabe DP. Needle aspiration biopsy of major salivary gland tumors: its value. Cancer 1986;57:554557. 15. Zajicek J. Aspiration biopsy cytology, part I: cytol-
ogy of supra-diaphragmatic organs. In: (series) Monographs in clinical cytology, vol. 4. New York: S. Karger, 1974~1-5. 16. Lindberg LG, Akeramn M. Aspiration cytology of sali-
vary gland tumors: diagnostic experience from six years of routine laboratory work. Laryngoscope 1976;86:584594. 17. Qizilbash AH, Sianol J, Young JE, et al. Fine needle as-
piration biopsy cytology of major salivary glands. Actu Cytol 1985;29:503-512. 18. Layfield LJ, Tan P, Glasgow BJ. Fine-needle aspiration of salivary gland lesions. Arch Pathol Lab Med 1987; 111:346-353. 19. Bono A, Chiesa F, Sala L, et al. Fine-needle aspiration biopsy in parotid masses. Tumori 1983;69:417421. 20. Persson PS, Zettergren L. Cytologic diagnosis of salivary gland tumors by aspiration biopsy. Actu Cytol 1973; 17~351-354.
HEAD & NECK
NovemberlDecember 1992
487
