Cytopathology 1990,1,25-33

Fine needle aspiration cytology of salivary gland lesions J. A. Y O U N G , L. A. S M A L L M A N , H. T H O M P S O N , D. W. PROOPS* A N D A. P. J O H N S O N * Department of Pathology, University of Birmingham and *Department of Otorhinolaryngology, Queen Elizabeth Hospital, Birmingham Accepted for publication 29 September 1989

YOUNG J. A., SMALLMAN L. A., THOMPSON H., PROOPS D.

w.

AND JOHNSON A. P.

(1990)

Cytopathology 1,25-33

Fine needle aspiration cytology of salivary gland lesions Eighty-eight fine needle aspirates from 79 salivary gland lesions in 77 patients were examined. The overall diagnostic sensitivity was 84% and the specificity 98.41%. When the 14 unsatisfactory specimens were excluded the sensitivity rose to 95.45%. Correct identification of the disease process was possible in nearly 80% of cases with a final benign diagnosis. The histological tumour type was correctly predicted in 75% of the malignancies. In the others the cytological diagnosis was anaplastic malignant neoplasm. Keywords: cytodiagnosis, fine needle aspiration cytology, salivary gland disease, salivary gland neoplasms

INTRODUCTION The clinical assessment of lesions in the salivary glands may be difficult and biopsy, whether open or Tru-cut, is unpopular due to the fear of complications. The value of fine needle aspiration cytology (FNAC) in the management of patients attending the Department of Otorhinolaryngology has been demonstrated by us in a previous report'. Aspirates from the salivary glands represented 21.13% of the cases described in this earlier paper and proved to be one of the most rewarding areas of the study. Our further experience in this field is now presented. MATERIALS A N D METHODS A total of 88 FNAC specimens from 79 separate salivary gland masses were obtained from 77 patients. Two patients had double lesions, one in the right and left parotid and the other in the parotid and submandibular glands. Nine repeat aspirations from the same site were Correspondence: Dr J. A. Young, Department of Pathology, Medical School, Birmingham B15 2TJ.

26 J . A. Younget al. Table 1. Anatomical site of salivary gland lesions and number of aspirates

Site

No. of lesions

No. of aspirates

Parotid Submandibular Hard palate Sublingual

41 27 3

2

28 3 2

Total

19

88

55

collected, either because the first attempt was unsatisfactory or because no specific diagnosis was possible. The anatomical distribution of the lesions and numbers of aspirations from each site are shown in Table 1. Forty-one of the patients were female and 36 male. The age range was 18 to 88 years with a mean 55 years. Aspiration was performed by the surgeons (D.W.P. and A.P.J.) at the bedside or in the outpatient clinic using the method previously described'.2. Two to four slides were prepared from each specimen. Half were wet fixed in 95% alcohol and stained by the Papanicolaou (Pap) technique and the remainder air-dried and stained with May-Grunwald-Giemsa (MGG). The cytopathology was reported by J.A.Y. and H.T. In addition to assessing the adequacy of the material and whether the lesion was benign or malignant, an attempt was made in each case to establish a precise diagnosis and to type all tumours. FNAC results were then compared with the final diagnoses. These were based on histology (L.A.S.) in 62 cases and on clinical follow-up for at least one year in the remaining 15 patients. RESULTS Aspiration was well tolerated by all patients and there were no complications. The overall results expressed in terms of malignant, benign and unsatisfactory are shown in Table 2. The unsatisfactory results are grouped with benign as showing 'no identifiable evidence of malignancy' giving sensitivity of 84% and specificity of 98.41YO.The final diagnosis in the 14 (15.91%) cases with unsatisfactory aspirates was malignant in three and benign in 11. When these are excluded and only the 74 (84.09%) satisfactory specimens are considered the sensitivity rises to 95.45%. The positive predictive value was also 95.45% and the negative predictive value 98.08%. The one false positive FNAC diagnosis occurred in an elderly patient with a hard parotid mass. This subsided spontaneously after a calculus was extruded and was considered clinically to be due to chronic sialadenitis. No histology was obtained. The false negative report among the satisfactory specimens was from a well-differentiated adenoid cystic carcinoma in the palate which was mistaken cytologically as a pleomorphic adenoma. The initial excision biopsy from this tumour was also interpreted as pleomorphic adenoma and it was not until multiple sections of the resected mass were examined that the true diagnosis of adenoid cystic carcinoma was made. In another example of adenoid cystic carcinoma of the submandibular gland, correctly typed by FNAC, a prior Tru-cut biopsy had been reported as pleomorphic

FNAC of salivary gland lesions 27 Table 2. Overall results of FNAC

Final diagnosis FNAC diagnosis Malignancy Benign

Unsatisfactory Total

Malignant Benign Sensitivity 22 52

21

1

1

51

14

3

11

88

25

63

84%

Specificity

98.41%

adenoma and the correct histological diagnosis was not made until the excision specimen was received. The final diagnosis and the full FNAC results for benign (neoplastic and non-neoplastic) and malignant cases are shown in Tables 3 and 4. Among the benign lesions the final diagnosis was non-specific in eight. If these and the unsatisfactory specimens are excluded FNAC provided the correct identification of the precise disease process in 35 of 44 (79.55%) cases with a final benign diagnosis. The cytological diagnosis of chronic sialadenitis proved to be the most difficult aspect of non-neoplastic benign disease. In only two out of six cases was the condition recognizable from the smear (Figure I). In two further examples the specimens were unsatisfactory and this group included the one false positive result discussed previously. Two of three cases of myoepithelial sialadenitis (MESA) were identified (Figure 2). When the specimen was satisfactory, typing by FNAC of benign salivary gland neoplasms was accurate. Eighteen of 21 (85.71 YO)pleomorphic adenomas (Figure 3) were correctly classified. In one the aspirate was unsatisfactory and in the remaining two a non-specific benign diagnosis was made. Seven of 10 adenolymphomas were recognized (Figure 4).In the other three, FNAC was non-specific in two and unsatisfactory in one. All the malignant tumours from which satisfactory specimens were obtained, were correctly identified as malignant by FNAC except for the adenoid cystic carcinoma of the palate mistaken for a pleomorphic adenoma. The histological tumour classification was predicted in 16 of the 20 (75%) malignant aspirates, while the remainder were reported as anaplastic carcinomas (Figure 5) one of which was indeed classified as anaplastic carcinoma following full histological investigation. The final diagnosis was squamouscell carcinoma in eight patients (Figure 6). In six of these the lesion was considered metastatic to intraparotid lymph nodes as the patients were known to have primary squamous cell carcinoma elsewhere. The remaining two cases were confirmed as a primary squamous cell carcinoma of salivary gland origin. The two cases of adenocarcinoma were also considered primary neoplasms of salivary glands after extensive chemical and histopathological investigation excluded an alternative primary site. DISCUSSION Evaluation of the diagnosis of salivary gland lesions by FNAC was largely pioneered by workers at the Karolinska Institute, Sto~kholm.’~FNAC of masses in the salivary glands remains one of the least utilized applications of the technique outside Scandinavia despite the need for a safe, efficient method of diagnosing these tumours. Clinical assessment alone is

28 J. A. Young et al. Table 3. Results of FNAC in comparison with final diagnosis: benign cases Final diagnosis Normal Acute sialadenitis Chronic sialadenitis

MESA Amyloidosis Pleomorphic adenoma

Adenolymphoma

FNAC diagnosis

1 3 6

3

1 21

10

Lipoma Lymphadenitis (intraparotid lymph node)

2

Non-specific benign

8

Total

1

56 lesions

Normal Acute sialadenitis Chronic sialadenitis Non-specific benign Carcinoma Unsatisfactory MESA Non-specific benign* Non-specific benign Pleomorphic adenoma Non-specific benign Unsatisfactory Adenolymphoma Non-specific benign Unsatisfactory Lipoma Lymphadenitis Non-specific benign Non-specific benign Unsatisfactory

1 3 2 1 1

2 2 1 1 18

2 1

I 2 1 1 1 1

2 6

56 lesions

*2nd aspirate, 1st unsatisfactory

unreliable and Tru-cut biopsy is not appropriate due to the risk of fistulae and haemorrhage. The main objection to aspirating salivary gland tumours has been the danger of seeding neoplastic cells along the needle track. However, extensive clinical research has been carried out to prove its safety. Frable' searched the literature for reports on the spreading or seeding of tumour cells by the needle and found only two cases both of which followed Tru-cut biopsy. No case of seeding following FNAC of salivary gland tumours was reported. Engzell et a/." reported a 10-15 year follow-up of 157 patients with pleomorphic adenoma of the major salivary glands on whom FNAC had been performed. There was no evidence of recurrence or local extension of tumour that could be attributed to the diagnostic procedure. The overall level of diagnostic specificity (98.41 %) in this study was high and demonstrated the reliability of a malignant cytological diagnosis. There was only one false positive result among the 56 benign lesions. This was a parotid mass clinically considered to be chronic sialadenitis,from which no histology was obtained. The overall diagnostic sensitivity was 84%. However, in three of the four malignant tumours not identified, the FNAC specimens were unsatisfactory and when these cases are excluded the sensitivity rises to 98.45%. As with FNAC from other sites the cytodiagnosis should always be considered

FNAC of salivary gland lesions 29 Table 4. Results of FNAC in comparison with final diagnosis: malignant cases Final diagnosis

FNAC diagnosis

Adenoid cystic carcinoma

8

Squamous cell carcinoma

8

Adenocarcinoma Mucoepidermoid carcinoma Carcinoma ex-pleomorphic adenoma Anaplastic carcinoma Lymphoma Leu kaemia

2

Total

23 lesions

1

Adenoid cystic carcinoma Pleomorphic adenoma Squamous cell carcinoma Anaplastic carcinoma Unsatisfactory Adenocarcinoma Anaplastic carcinoma

7 1 5 1

2 2 1 1

1

Unsatisfactory Anaplastic carcinoma Lymphoma

I

Anaplastic malignant cells

1

1

1

1 1

23 lesions

Figure 1. Chronic sialadenitis. Groups of epithelial cells against a background of blood and debris. MGG. x 400.

in the context of the clinical findings and unsatisfactory or non-diagnostic specimens discounted appropriately. Layfield et ul." in a review article, estimate that aspirates from over 3000 cases of salivary gland disease are now described in 36 published studies. However, it is difficult to extract the true diagnostic sensitivity and specificity from the figures in many of these papers, either because so often salivary gland tumours are amalgamated with those elsewhere in the head and neck or because the inclusion of unsatisfactory or suspicious results makes the data difficult to interpret. Early reports" contained a higher than expected false positive (or falsely suspicious) rate due largely to misinterpretation of cells aspirated from pleomorphic

30 J . A. Younget al.

Figure 2. Myoepithelial sialadenitis (MESA). Streaks of lymphocytes and plasma cells. MGG. x560.

Figure 3. Pleomorphic adenoma. Epithelial cells and mesenchymal elements (myxoid material) closely intermingled. Mesenchymal elements not shown. MGG. x 560.

Figure 4. Adenolymphoma. Cluster of oncocytic cells with scattered lymphocytes and a background of amorphous debris. MGG. x 400.

FNAC of salivary gland lesions 3 1

Figure 5. Anaplastic malignant cells. Loose collection of large undifferentiatedcells. MGG. x 560.

Figure 6. Squamous cell carcinoma. Keratinized malignant squamous cells from metastaticcarcinoma in an intraparotid lymph node.

MGG. x400.

adenomas. Recent papers which critically assess results show a false positive rate of 0°hi3 to 4.7%14.Similarly early studiesdemonstrated a high false negative rate of up to 37%I2*I5while in some later series no false negative diagnoses had been givenI6. Layfield et a/." found false negative rates to be generally under 10%. The groups with the most accurate results achieved a very high level of satisfactory aspiralthough in some instances more ates most having less than 3% inadequate than one attempt at aspiration was required to obtain this satisfactory material. The overall unsatisfactory rate in our study was 15.91 %. Many of the failures were from the early cases in the series and with increased experience in aspiration the incidence of unsatisfactory specimens has declined. For FNAC of salivary glands to be clinically efficacious a more precise diagnosis than benign or malignant is required. A mass may be due to inflammation, cyst or benign or malignant neoplasm of diverse variety. Abnormal intraparotid lymph nodes may also lead to salivary gland enlargement and cause particular problems of interpretation for the cytopathologist. If perinodal ductal or acinar cells are aspirated as well as lymphoid cells it may be

32 J . A. Young et al.

Figure 7. Adenoid cystic carcinoma. Groups of malignant cells surrounding globules of basement membrane material. MGG. x 400.

difficult to distinguish the smear pattern from that found in FNAC of a lymphoid-rich salivary gland les:on such as MESA or adenolymphoma. In this study, as shown in Table 3, it was possible to make a precise diagnosis in 79.55% of benign cases with a specific final diagnosis. Typing of benign neoplasms including pleomorphic adenoma, adenolymphoma and lipoma was satisfactory. Identification of non-neoplastic lesions with a specific final diagnosis-especially chronic sialadenitis'* was more difficult. Adenoid cystic carcinoma was the most common primary malignant tumour (Table 4). Apart from the case in the palate all seven tumours were correctly typed. The neoplastic cells are small and closely packed and the large globules of basement membrane material, when numerous, are a distinctive feature (Figure 7). However, pleomorphic adenomas occasionally contain areas indistinguishable from well-differentiated adenoid cystic carcinoma and a few globules may be found in aspirates from such a lesion. Poorly differentiated squamous cell carcinoma and mucoepidermoid carcinoma are particularly difficult to distinguish. The finding of differentiated malignant squamous cells (Figure 6) also raises problems of interpretation. These may arise from the squamoid component of a mucoepidermoid carcinoma, from a primary squamous cell carcinoma or most commonly from metastatic carcinoma in intraparotid lymph nodes. Although both MESA and lymphoma were correctly diagnosed in this study the distinction between benign and neoplastic lymphoproliferation must always be approached with caution when only limited cellular material is available. There was one case of leukaemic infiltration in this series. This was reported simply as 'anaplastic malignant cells' as the patient was known to have leukaemia, carcinoma of the colon and melanoma and the aspirated material was insufficient for special stains. FNAC of salivary gland masses is rapid, safe and relatively painless. With experience in interpretation, providing the specimen is satisfactory, the method is sensitive and specific in establishing the presence of malignancy and in many cases can provide a precise diagnosis. ACKNOWLEDGEMENTS We wish to thank Mr W. P. Cuthbertson and Ms L. Grosvenor for technical assistance, Mrs Elaine Haywood for help with photography and Miss A. J. Wright for typing the manuscript.

FNAC of salivary gland lesions 33

REFERENCES I Smallman LA, Young JA, Oates J, Proops DW, Johnson AP. Fine needle aspiration cytology in the management of ENT patients. J Laryngol Otol 1988; 102 909-13. 2 Young JA. Fine needle aspiration of salivary glands. Ear Nose Throat J 1989; 68: 12&9. 3 Maves P, Eneroth C-M, Franzen S , Moberger G, Zajicek J. Aspiration biopsy of salivary gland tumors. I. Correlation of cytologic reports from 652 aspiration biopsies with clinical and histologic findings. Acta Otolaryngol (Stockh) 1964; S8: 472-84. 4 Eneroth C-M, Zajicek J. Aspiration biopsy of salivary gland tumors. 11. Morphologic studies on smears and histologic sections from oncocytic tumours (45 cases of papillary cystadenoma lymphomatosum and 4 cases of oncocytoma). Acta Cyrol1965; 9 355-6 I . 5 Eneroth C-M, Zajicek J. Aspiration biopsy of salivary gland tumors. 111. Morphologic studies on smears and histologic sections from 368 mixed tumours. Acta Cytol1966; 1 0 440-54. 6 Eneroth C-M, 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: 5943. 7 Eneroth C-M, Zajicek J. Aspiration biopsy of salivary gland tumors. V. Morphologic investigations on smears and histologic sections of acinic cell carcinoma. Acta Radiol (Stockh) 1971; 31O(S~ppl):85-93. 8 Zajicek J, Eneroth C-M, Jakobsson P. Aspiration biopsy of salivary gland tumors. V1. Morphologic investigation on smears and histologic sections of

24 cases with mucoepidermoid carcinoma. Acta Cyroll976; 2 0 35-41. 9 Frable WJ. Thin needle aspiration biopsy. Am J Clin Pathol1976; 6 5 68-181. 10 Engzell V, Esposti DL, Rubio C, Sigurdson A, Zajick J. Investigation of tumour spread in connection with aspiration biopsy. Acta Radiol (Therapeutic) 1971; 1 0 385-98. 11 Layfield LJ, Tan P. Glasgow BJ. Fine needle aspiration of salivary gland lesions: comparison with frozen sections and histologic findings. Arch Pathol Lab Med 1987; iii: 3 4 6 5 3 . I2 Eneroth C-M, Franzen S , Zajicek J. Cytologic diagnosis on aspirates from lo00 salivary gland tumors. Acta Otolaryngol (Stockh) 1967; 224(S~ppl):168-71. 13 Qizilbash AH, Sianos J, Young JEM, Archibald SD. Fine needle aspiration biopsy cytology of major salivary glands. Acta Cvtol 1985; 2 9 50312. 14 ODwyer P, Farrar WB, James AG, Finkelmeler W, McCabe DP. Needle aspiration biopsy of major salivary gland tumours: its value. Cancer 1986; 57: 554-7. 15 Zajicek J, Eneroth C-M. Cytological diagnosis of salivary gland carcinomata from aspiration biopsy smears. Acta Orolaryngoll970; 263: 183-5. 16 Webb AJ. Cytological diagnosis of salivary gland lesions in adult and paediatric surgical patients. Acta Cytoll973; 17:51-8. 17 Sismanis A, Merriam JM, Kline TS et a/. Diagnosis of salivary gland tumours by fine needle aspiration biopsy. Head Neck Surg 1981; 3: 482-9. 18 Kondratowicz GM, Smallman LA, Morgan DA. A clinicopathological study of myoepithelial sialadenitis and chronic sialadenitis/sialolithiasis.J Clin Pathol1988; 41: 403-9.

Fine needle aspiration cytology of salivary gland lesions.

Eighty-eight fine needle aspirates from 79 salivary gland lesions in 77 patients were examined. The overall diagnostic sensitivity was 84% and the spe...
1MB Sizes 0 Downloads 0 Views