Oncocytoma

Context.-We use data from the College of American Pathologists lnterlaboratory Comparison Program in Nongynecologic Cytology to identify common diagnostic errors in salivary gland fine-needle aspiration (FNA).

Objective.-To identify salivary gland FNA cases with poor performance characteristics in the Nongynecologic Cytology Program surveys, so that the most common diagnostic pitfalls can be avoided.

Design.-A retrospective review of the College of American Pathologists Nongynecologic Cytology Program's cumulative data from 1999 to 2003 revealed the most common false-positive and false-negative interpretations on FNA for common salivary gland lesions. Slides that performed poorly were then reviewed to identify the cytologic characteristics that may have contributed to their poor performance.

Results.-A total of 6249 participant responses with general interpretations of benign (n = 4642) or malignant (n = 1607) were reviewed. The sensitivity and specificity of the participant responses for correctly interpreting the cases as benign or malignant were 73% and 91%, respectively. Benign cases with the highest false-positive rates were monomorphic adenoma (53% false-positive), intraparotid lymph node (36%), oncocytoma (18%), and granulomatous sialadenitis (10%). Malignant cases with the highest false-negative rates were lymphoma (57%), acinic cell carcinoma (49%), low-grade mucoepidermoid carcinoma (43%), and adenoid cystic carcinoma (33%). Selected review of the most discordant individual cases revealed possible explanations for some of the interpretative errors.

Conclusions.-These data confirm the difficulty associated with interpretation of salivary gland FNA specimens. Cytologists should be aware of the potential false-positive and false-negative interpretations that can occur in FNAs from this organ site in order to minimize the possibility of diagnostic errors.

(Arch Pathol Lab Med. 2005;129:26-31)

The Interlaboratory Comparison Program in Nongynecologic Cytopathology (NGC) was inaugurated by the College of American Pathologists (CAP) in 1997. Most participating laboratories are in the United States, although a small number of international laboratories also subscribe. Participating NGC Program laboratories receive quarterly mailings of 5 glass slides with accompanying clinical histories. The slides are contributed to the program by members of the CAP Cytopathology Resource Committee and are reviewed at several screening sessions for consensus prior to inclusion in the program. All of the Cytopathology Resource Committee pathologists are certified by the American Board of Pathology and hold subspecialty certification in Cytopathology. Many, if not most, of the Cytopathology Resource Committee members are nationally and internationally recognized experts in fine-needle aspiration (FNA) cytology. Submitted cases must be validated by 2 Cytopathology Resource Committee cytopathologists, who agree on the general and specific interpretation and who also agree that the slide is representative and has no technical inadequacies that might inhibit an accurate assessment. An interpretative menu that uses a bubble answer format is provided to the participating laboratories with the glass slides. The participants' responses are tabulated by the CAP, which then provides a list of the reference diagnoses to the participating laboratories for education and self-assessment. Individual participants use the CAP NGC Program as an educational tool as well as a measure of their performance as compared with other laboratories. However, because slides circulate extensively, most slides have a unique accrued slide profile, allowing evaluation of the characteristics of each slide in the program.

The data generated by the NGC survey provide a valuable source of information about which types of cases present the most challenging interpretation problems for cytologists from a wide range of practice settings. In this study, we used data from the NGC Program to examine the accuracy of FNA interpretation for salivary gland lesions and to determine the most common sources of errors.

MATERIALS AND METHODS

Cumulative histories of all NGC Program participant responses for all salivary gland FNA cases from 1999 through 2003 were obtained through the CAP's "Scores" computer system. The cases were divided into general and reference diagnostic categories for comparison within category. The general categories were negative, positive, suspicious for malignancy, and unsatisfactory. The reference diagnoses available to participants are listed in Table 1.

In the first part of the study, the year-end data compiled for all of the reference interpretations were reviewed, and the overall accuracy, false-positive, and false-negative rates of the NGC Program participants were calculated. For the purposes of determining false-positive and false-negative rates, participant responses of "suspicious for malignancy" were considered "positive for malignancy." The most common interpretational errors were also tabulated.

In the second part of the study, data from the Scores system were used to identify individual slides that had a high level of discordance between the reference interpretation and the NGC Program participants' responses. Although many of these slides had not circulated enough times to permit rigorous statistical evaluation of their performance characteristics, many of them showed consistent, reproducible errors that made their analysis a potential source of information for identifying common pitfalls. A selective review of these slides was undertaken by 2 of the authors (J.H.H., D.C.W.), and an effort was made to determine the most likely causes for the interpretation errors.

RESULTS

The results of the 1999-2003 retrospective compilation of NGC Program participants' responses for salivary gland FNA diagnostic categories are presented in Tables 2 and 3.

Table 2 shows the tabulated responses for the slides with an established general diagnosis of benign. There were a total of 4642 responses, 4254 of which carried a correct general interpretation of benign, resulting in a specificity of 91%. Three hundred eighty-eight of the cases were incorrectly interpreted as malignant, resulting in a false-positive rate of 8%. The highest false-positive rates were seen in cases of monomorphic adenoma (53% falsepositive rate), intraparotid lymph node (36%), oncocytoma (18%), and granulomatous sialadenitis (10%). Pleomorphic adenoma and Warthin tumor, which constituted the large majority of the benign cases, were associated with false-positive rates of 8% to 9%. The most common false-positive interpretation for pleomorphic adenoma was adenoid cystic carcinoma, and the most common falsepositive interpretation for Warthin tumor was lymphoma. The overall accuracy of the CAP NGC Program participants for making the correct specific reference interpretation was 76%.

Table 3 shows the tabulated responses for the slides with an established general interpretation of malignant. There were a total of 1607 responses, 1096 of which carried a correct general diagnosis of malignant, resulting in a sensitivity of 68%. Five hundred eleven cases were incorrectly diagnosed as benign, resulting in a false-negative rate of 32%. The highest false-negative rates were seen in cases of lymphoma (57% false-negative rate), acinic cell carcinoma (49%), low-grade mucoepidermoid carcinoma (43%), and adenoid cystic carcinoma (33%). The overall accuracy of the CAP NGC Program participants for making the correct specific reference interpretation was 48%.

The second part of the study involved the selective review of individual cases that showed a high degree of discordance between the reference diagnosis and the NGC Program participants' interpretations. Most of these cases had not circulated enough to generate sufficient participant responses for meaningful statistical analysis of their performance, but they were identified by the Scores system as highly discordant cases. Because these highly discordant cases represented a unique subset of cases that might allow insight into potential interpretative pitfalls, they were reviewed in an effort to identify the possible sources of the errors. Some of these cases were found to demonstrate low cellularity or poor stain quality, and these factors probably contributed to their poor performance. However, other discordant cases had good cellularity and were technically adequate. Some examples of these latter cases are shown in Figures 1 and 2.

Figure 1, A, shows a Diff-Quik-stained case of pleomorphic adenoma, incorrectly interpreted as acinic cell carcinoma. In our analysis of this case, we concluded that the most likely reason for this diagnostic error was the failure to recognize the stromal component of the pleomorphic adenoma, which was not very prominent in this case. The scantness of the stroma, when coupled with focal areas of pseudo-acinar architecture, probably contributed to the poor performance of this case. Had the stromal component been identified, confusion with acinic cell carcinoma would be unlikely, given the bland cytologic features of the cells.

Figure 1, B, shows a Papanicolaou-stained case of acinic cell carcinoma, incorrectly diagnosed as normal salivary gland. Upon reviewing this case, we felt that the most likely reason for this error was the failure to appreciate the cellularity of the lesion, which was greater than that typically seen in nonneoplastic processes. Additionally, some participants may not have appreciated the architecture of the acinar structures, which were more disorganized and discohesive than those of normal salivary gland.

Figure 1, C, shows a Papanicolaou-stained case of adenoid cystic carcinoma, incorrectly interpreted as lymphoma by some respondents. This error almost certainly resulted from failure to recognize the stromal component of the adenoid cystic carcinoma. Had the stroma been appreciated, it is unlikely that a diagnosis of lymphoma would have been entertained. However, it is easy to see how failure to recognize the stroma, when coupled with the bland monomorphic cytology of the cells, could lead to the incorrect impression of an atypical lymphoproliferative process.

Figure 1, D, shows a Papanicolaou-stained case of Warthin tumor, incorrectly interpreted as lymphoma. This is a well-recognized pitfall that usually results from failure to recognize the oncocytic component of the tumor, particularly in cases in which the lymphoid component predominates.

Figure 1, E, shows a Papanicolaou-stained case of adenoid cystic carcinoma, incorrectly interpreted as monomorphic adenoma. Again, this error probably reflects failure to recognize the stromal component of adenoid cystic carcinoma on the Papanicolaou stain. The false-negative diagnosis of adenoid cystic carcinoma may also be avoided by careful attention to the nuclear features of the tumor; the presence of visible nucleoli and any abnormality of the chromatin favor a diagnosis of adenoid cystic carcinoma over monomorphic adenoma, although these helpful cytologie features are not present in every case.

Figure 1, F, shows a Papanicolaou-stained case of metastatic small cell carcinoma, incorrectly interpreted as lymphoma. This is another well-described pitfall, but it might have been avoided in this case by recognizing the absence of lymphoglandular bodies.

Figure 2, A, shows a case of mucoepidermoid carcinoma that was incorrectly interpreted as Warthin tumor. It is likely that the abundant dirty background material, which is commonly seen in Warthin tumor, and the eosinophilic tumor cells, which may have been misinterpreted as oncocytes, led to this error in classification.

Figure 2, B, shows a case of oncocytoma that was incorrectly interpreted as pleomorphic adenoma by all respondents. This interpretative error may have been the result of mistaking the vascularity of the oncocytoma for the stranded stroma of pleomorphic adenoma.

Figure 2, C, shows a case of reactive lymphoid hyperplasia in an intraparotid lymph node that was incorrectly interpreted as lymphoma by the majority of respondents. Although low-grade lymphoma enters the differential diagnosis of any reactive process, the polymorphous population of lymphocytes in this case favors a benign process. This problematic case also underscores the importance of ancillary immunophenotyping studies in the evaluation of any cytologically suspicious lymph node. Participants in the NGC Program were not supplied with immunophenotyping results on this case, but such studies would certainly have been useful in excluding a low-grade lymphoproliferative disorder.

Figure 2, D, shows a case of pleomorphic adenoma that NGC Program participants incorrectly interpreted as low-grade mucoepidermoid carcinoma. In this case, it is likely that the stroma of the pleomorphic adenoma may have been misinterpreted as mucin, thereby leading to the erroneous conclusion that this lesion represents a low-grade mucoepidermoid carcinoma.

COMMENT

Salivary gland FNAs are a common specimen in most pathology practices and present difficult interpretation challenges. Several large published series have documented the accuracy and limitations of salivary gland FNA. The overall accuracy has been reported to be 87% to 100% in distinguishing benign from malignant lesions; FNA also has a reported sensitivity of 87% to 100% and a specificity of 90% to 100%.1-8 Most of these series have been generated at large academic centers. To our knowledge, our study is unique in that the CAP NGC Program assesses salivary gland FNA accuracy across a diverse group of practicing pathologists, including academic centers, commercial laboratories, and large and small private practice settings; therefore, these data may more accurately reflect the diversity of practice conditions. Moreover, because the same slides circulate among numerous participants, the participants' responses can be tabulated and used to identify particularly problematic individual cases that provide insight into common diagnostic dilemmas and pitfalls.

Our 5-year review of NGC survey results revealed a 73% sensitivity for making the correct general diagnosis of malignant, a 91% specificity for making the correct general diagnosis of benign, and an overall diagnostic accuracy of 48% for making the correct specific reference diagnosis. These data are similar to those observed in the reported large academic series. The benign lesions that were most often misdiagnosed as malignant were monomorphic adenoma (53% false-positive rate), intraparotid lymph node (36%), oncocytoma (18%), and granulomatous sialadenitis (10%). The malignant lesions that were most often misdiagnosed as benign were lymphoma (57%), acinic cell carcinoma (49%), low-grade mucoepidermoid carcinoma (43%), and adenoid cystic carcinoma (33%).

Our selected review of the most discordant cases provided some insight into the sources of possible interpretation errors. Some of the discordant cases represented well-described diagnostic pitfalls, such as the false-negative interpretation of acinic cell carcinoma as normal salivary gland9 or the false-positive interpretation of Warthin tumor as lymphoma.10 However, some of the most discordant cases were not expected. Although we can only speculate as to the reason why some well-validated, technically excellent slides did not perform well, some of the examples of discordant cases depicted in Figure 1 suggest that failure to recognize the diagnostically helpful stromal component on Papanicolaou-stained slides may have been an important factor. This interpretation underscores the importance of using both a Romanowsky-type stain (such as Diff-Quik) and the Papanicolaou stain when examining FNA material from salivary gland, because stromal components may be more readily identified on the Romanowsky stain. Many of the false-positive and false-negative interpretations of lymphoma probably resulted from the lack of any accompanying flow cytometry immunophenotyping data with the cases. The poor performance of these cases highlights the importance of immunophenoryping studies to the interpretation of lymphoid lesions, particularly low-grade lymphoproliferative processes.

Using a 5-year cumulative slide history from the CAP NGC database, we examined the diagnostic accuracy of the NGC Program participants for salivary gland FNA. This review included 6249 responses for a wide variety of benign and malignant salivary gland lesions. The data from our review demonstrated that salivary gland FNA is 73% sensitive and 91% specific for distinguishing between benign and malignant lesions. The overall accuracy of FNA for correctly identifying the correct specific reference interpretation was 48%. Slide review from cases of high discordance suggested that some recurrent interpretations might be the result of technically suboptimal slides. Among the technically adequate slides, some of the most discordant slides represented classic diagnostic pitfalls,11-19 such as acinic cell carcinoma versus normal salivary gland, lymphoma versus Warthin tumor, and adenoid cystic carcinoma versus monomorphic adenoma. Some discordant cases appeared to occur as a result of participants' not recognizing the stromal component of pleomorphic adenoma or adenoid cystic carcinoma on Papanicolaoustained slides; these cases underscored the importance of using a Romanowsky-type stain, which is superior to the Papanicolaou stain for assessing stromal elements, on salivary gland FNAs. The study also demonstrated the importance of ancillary immunophenotyping studies for classifying lymphoid lesions. Awareness of some of these pitfalls and application of classic criteria may help improve performance characteristics of salivary gland FNA specimens in the CAP NGC and in daily practice.

We thank Jennifer Haja, College of American Pathologists Non-gynecologic Cytology Program committee staff member, for her help with data procurement and slide handling.

References

1. Schindler S, Nayar R, Outra J, Bedrossian CW. Diagnostic challenges in aspiration cytology of the salivary glands. Semin Diagn Pathol. 2001;18:124-146.

2. Stewart CJ, MacKenzie K, McCarry GW, Mowat A. Fine-needle aspiration cytology of salivary gland: a review of 341 cases. Diagn Cytopathol. 2000;22: 139-146.

3. Cajulis RS, Gokaslan ST, Yu GH, Frias-Hidvegi D. Fine needle aspiration biopsy of the salivary glands: a five-year experience with emphasis on diagnostic pitfalls. Acte Cytol. 1997;41:1412-1420.

4. Shintani S, Matsuura H, Hasegawa Y. Fine needle aspiration of salivary gland tumors. Int J Oral Maxillofac Surg. 1997;26:284-286.

5. Zarka MA. Fine-needle aspiration of the salivary glands. Pathology. 1996;4: 287-318.

6. Orell SR. Diagnostic difficulties in the interpretation of fine needle aspirates of salivary gland lesions: the problem revisited. Cytopathology. 1995;6:285-300.

7. MacLeod CB, Frable WJ. Fine-needle aspiration biopsy of the salivary gland: problem cases. Diagn Cytopathol. 1993;9:216-224.

8. el Hag IA, Chiedozi LC, al Reyees FA, Kollur SM. Fine needle aspiration cytology of head and neck masses: seven years' experience in a secondary care hospital. Acte Cytol. 2003;47:387-392.

9. Nagel H, Laskawi R, Buter JJ, Schroder M, Chilla R, Droese M. Cytologic diagnosis of acinic-cell carcinoma of salivary glands. Diagn Cytopathol. 1997; 16:402-412.

10. Flezar M, Pogacnik A. Warthin's tumour: unusual vs. common morphological findings in fine needle aspiration biopsies. Cytopathology. 2002:13:232-241.

11. Chhieng DC, Cangiarella JF, Cohen JM. Fine-needle aspiration cytology of lymphoproliferative lesions involving the major salivary glands. Am J Clin Pathot. 2000;113:563-571.

12. Stewart CJ, Jackson R, Farquharson M, Richmond J. Fine-needle aspiration cytology of extranodal lymphoma. Diagn Cytopathol. 1998;19:260-266.

13. Lussier C, Klijanienko J, Vielh P. Fine-needle aspiration of metastatic non-lymphomatous tumors to the major salivary glands: a clinicopathologic study of 40 cases cytologically diagnosed and histologically correlated. Cancer. 2000;90: 350-356.

14. MacCallum PL, Lampe HB, Cramer H, Matthews TW. Fine-needle aspiration cytology of lymphoid lesions of the salivary gland: a review of 35 cases. J Otolaryngol. 1996;25:300-304.

15. Cohen MB, Fisher PE, Holly EA, Ljung BM, Lowhagen T, Bottles K. Fine needle aspiration biopsy diagnosis of mucoepidermoid carcinoma: statistical analysis. Acte Cytol. 1990;34:43-49.

16. Elliott JN, Oertel YC. Lymphoepithelial cysts of the salivary glands: histologicand cytologic features. Am J Clin Pathol. 1990;93:39-43.

17. Mair S, Phillips JI, Cohen R. Small cell undifferentiated carcinoma of the parotid gland: cytologic, histologic, immunohistochemical and ultrastructural features of a neuroendocrine variant. Acta Cytol. 1989;33:1 64-168.

18. Layfield LJ. Fine needle aspiration cytology of a trabecular adenoma of the parotid gland. Acta Cytol. 1985;29:999-1002.

19. 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;111:346-353.

Jonathan H. Hughes, MD, PhD; Emily E. Volk, MD; David C. Wilbur, MD; for the Cytopathology Resource Committee, College of American Pathologists

Accepted for publication September 9, 2004.

From Laboratory Medicine Consultants, Ltd, Las Vegas, Nev (Dr Hughes); the Department of Pathology, William Beaumont Hospital, Troy, Mich (Dr Volk); and the Department of Pathology, Massachusetts General Hospital, Boston (DrWilbur).

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Jonathan H. Hughes, MD, PhD, Laboratory Medicine Consultants, Ltd, 3059 S Maryland Pkwy, #100, Las Vegas, NV 89109-2201 (e-mail: jhughes627@aol.com).

Copyright College of American Pathologists Jan 2005
Provided by ProQuest Information and Learning Company. All rights Reserved