Journal of Cytology

: 2020  |  Volume : 37  |  Issue : 1  |  Page : 18--25

Evaluation of accuracy of milan system for reporting salivary gland cytology: Review of morphology and diagnostic challenges in each category

K Amita, HB Rakshitha, Avinash Singh, S Vijay Shankar 
 Department of Pathology, Adichunchanagiri Institute of Medical Sciences, Adichunchanagiri University, Mandya, Karnataka, India

Correspondence Address:
Dr. H B Rakshitha
Department of Pathology, Adichunchanagiri Institute of Medical Sciences, Adichunchanagiri University, BG Nagara, Nagamangala Taluk, Mandya district, Karnataka - 571448


Background: Fine-needle aspiration cytology (FNAC) in salivary gland lesions is challenging for the cytopathologists due to diverse morphological pattern and overlapping morphologic features which are responsible for the pitfalls. The present study was conducted to evaluate the efficacy of the Milan system in the diagnosis of salivary gland lesions and to discuss and review the morphology and diagnostic challenges in individual Milan categories. Materials and Methods: The study was a retrospective diagnostic analytical study in the department of pathology at a tertiary care hospital attached to medical college over a duration of 2 years. All the salivary gland FNAC cases were reviewed and divided into six categories as per the proposed Milan system for reporting salivary gland cytopathology. Histopathology correlation was performed wherever possible. Results: A total of 131 cases formed the study group. The number of cases in each category were: nondiagnostic 4.5%, nonneoplastic 51.9%, atypical lesions 0.76%, neoplastic category benign neoplasm 21.37%, salivary lesion of uncertain malignant potential 1.52%, suspicious category 2.29%, and malignant category 17.5%. The risk of malignancy for each categories were 6.25% (nonneoplastic), 100% (atypical), 3.3% (neoplastic), 0% (benign), 25% (salivary neoplasm of uncertain neoplastic potential), 100% (suspicious for malignancy), and 100% (malignant) categories. Sensitivity, specificity, positive predictive value, and negative predictive value of FNAC with application of Milan system was 89.4%, 100%, 100%, and 95.74%, respectively. Conclusion: The high efficacy of FNAC obtained in the present study, when Milan system was applied, confirms the usefulness of this scheme in reporting salivary gland lesions.

How to cite this article:
Amita K, Rakshitha H B, Singh A, Shankar S V. Evaluation of accuracy of milan system for reporting salivary gland cytology: Review of morphology and diagnostic challenges in each category.J Cytol 2020;37:18-25

How to cite this URL:
Amita K, Rakshitha H B, Singh A, Shankar S V. Evaluation of accuracy of milan system for reporting salivary gland cytology: Review of morphology and diagnostic challenges in each category. J Cytol [serial online] 2020 [cited 2020 Apr 7 ];37:18-25
Available from:

Full Text


The main purpose of fine-needle aspiration cytology (FNAC) in salivary gland lesions is to identify the lesion as salivary gland origin, classify them into benign or malignant, and further subtype them as low-grade or high-grade malignancy.[1] Several studies have emphasized that FNAC in salivary gland lesions is fraught with pitfalls and the reason for this is multifactorial.[2],[3],[4] In an attempt to increase the accuracy of the test, several other methods were endeavored like pattern-based analysis by few authors.[5],[6],[7] To bring uniformity in reporting, international experts in salivary gland cytopathology proposed the Milan system in 2015, a risk-based stratification system, similar to that of the Bethesda system for reporting cervical and thyroid cytology. This culminated in the atlas of the Milan system of reporting salivary gland cytopathology (MSRSGC) which was published in February 2018. Any new classification system needs to be validated for its practical applicability. Few studies have attempted to evaluate the efficacy of Milan system. The present study was conducted to evaluate its efficacy and to discuss and review the morphology and diagnostic challenges in individual Milan category.

 Materials and Methods

This was a 2-year retrospective diagnostic analytical study from January 2016 up to December 2017, in a tertiary care center. Ethical committee clearance was obtained from institutional ethical committee, the number being 131/2015. All the cases presenting to cytology clinic with salivary gland lesions for FNAC were included in the study. Recurrent lesions and noncooperative cases in which FNAC was not performed were excluded from the study.

Demographic data and radiologic findings were retrieved from the case records. Detailed local examination findings for the location were recorded from the case records. In our institute, the following protocols are routinely followed for all FNACs, which was also adopted in the present study. FNAC was done by the trained cytopathologist using 23 G needle. A maximum of two passes were performed. In the case of large swellings, aspiration was done from different areas to minimize sampling error. In case of cystic lesion, after complete evacuation, FNAC was done on any residual solid area. The fluid so aspirated was subjected to centrifugation and sediment smears prepared. Cell blocks were prepared from the sediment material and examined. In complex lesions (solid and cystic areas), FNAC was done under ultrasound guidance. The material was spread on slide and 50% were fixed in 90% ethanol for hematoxylin and eosin (H and E) stain and Papanicolaou stain and 50% were air-dried for Giemsa stain.

All the slides from each case were retrieved, the previous diagnosis was confirmed and then reviewed by two experienced cytopathologists. In cases of discrepancies, the diagnosis by senior cytopathologist was considered. All the cases were divided into six categories as per the proposed MSRSGC. Surgical pathology specimens in cases that underwent surgical excision underwent routine processing and cytohistopathology correlation was done.

Quantitative data were expressed as mean and standard deviation, percentages, and proportions. Considering histopathology as gold standard, sensitivity, specificity, positive predictive value, negative predictive value, diagnostic accuracy (DA) was calculated. The risk of malignancy (ROM) was determined for each category by dividing the number of cancer cases by number of cases with surgical follow-up.


All the cases presenting to the cytology clinic with salivary gland lesions for FNAC from January 2016 up to December 2017, that is, 136 cases were included in the study. Recurrent lesions which constituted two cases and three cases who were noncooperative for FNAC procedure in which FNAC was not performed were excluded from the study. Hence after final exclusion, the study included 131 cases. The mean age of patient was 48.3 years, the youngest patient was 9 years old and oldest was 83 years old. Males were more frequently affected with male to female ratio being 2.1:1. Parotid was the most frequently involved gland (94/131, 71.75%), followed by submandibular gland (33/131, 25.19%), palate (2/131, 1.52%), lacrimal gland (1/131, 0.76%), and sublingual gland (1/131, 0.76%).

The distribution of a total of 131 cases in various cytodiagnostic categories of Milan is shown in [Table 1]. [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8] [Table 2] shows cytohistopathology correlation.{Table 1}{Figure 1}{Figure 2}{Figure 3}{Figure 4}{Figure 5}{Figure 6}{Figure 7}{Figure 8}{Table 2}

Surgical follow-up was obtained in 48.85% (64/131) cases. Out of these 64 cases, concordance was noted in 44.27% (58/131) cases and the discordant diagnosis was rendered in 4.58% (6/131) cases. Among the discordant cases, one case of Warthin's tumor was underreported as chronic sialadenitis, while one case of chronic sialadenitis was wrongly interpreted as Warthin's tumor on FNAC. Subtyping error in tumor was noted in only one case, wherein an oncocytoma was misinterpreted as Warthin's tumor. [Table 3] Interestingly, there were two false negative (FN) and zero false-positive cases. The two FN cases were of low-grade mucoepidermoid carcinoma, each of which were wrongly reported as benign mucinous cyst and pleomorphic adenoma with atypia on cytology. [Figure 5], [Figure 6], [Figure 7] [Figure 8] Cellblock was performed in four cases in which fluid was aspirated. All the four cases yielded only cyst macrophages on FNAC. Ultrasound-guided FNAC was done in two cases. One case was a basal cell neoplasm wherein we had a difficulty in differentiating between basal cell adenoma and adenoid cystic carcinoma. Another case was a low-grade mucoepidermoid carcinoma which on initial FNAC yielded only cyst fluid, however, on ultrasound-guided FNAC we could accurately render a diagnosis of low-grade mucoepidermoid carcinoma which was confirmed on histopathology.{Table 3}

Sensitivity, specificity, positive predictive value, and negative predictive value of FNAC with the application of the Milan system for risk categorization were 89.4%, 100%, 100%, and 95.74%, respectively.

The ROM for the individual category is depicted in [Table 3]. It is recognized that the ROM has limitations, due to verification bias, which occurs based on the denominator used for calculating these values. DA for each category of the Milan system is depicted in [Table 2].


Though FNAC is the first-line investigation in the diagnosis of salivary gland lesions, it is fraught with pitfalls and the cause is multifactorial.[2],[3],[4] In order to overcome the pitfalls and increase the accuracy of FNAC, several modifications were embarked such as pattern-based analysis.[5],[6],[7] Christopher et al.[8] in 2015 proposed a simple risk-based classification model parallel to that available for thyroid and pancreas. Rossi et al.[9] classified salivary gland lesions into six categories and reported the ROM in benign and malignant lesions as 6.2% and 100%, respectively. With this background, MSRSGC was proposed to bring coherence in diagnosis and reporting of salivary gland lesions in a global front.[9] Meanwhile few studies attempted to apply the proposed Milan system in order to determine its practical applicability.[9],[10]

In the present study, 6 out of 131 cases (4.5%) showed no cellularity and were deemed as nondiagnostic. According to Griffith et al.,[5] the presence of epithelial cells in four high-power fields is required to call the smears as adequate. Those cases in which specific diagnosis was possible were also considered as adequate. In the present study, four cases showed only cyst macrophages and in the remaining two cases only hemorrhage was noted. Nondiagnostic rate as reported in the literature range from 5 to 10%. In one study, a high nondiagnostic rate of 28.2% was reported.[4],[5] The histopathologic follow-up of nondiagnostic cases shows a wide range from nonneoplastic to malignant lesions. Due to high rate of FN diagnosis in this category, Milan proposes to maintain an institutional nondiagnostic rate to below 10%.

In the present study, most common lesion in the nonneoplastic category was chronic sialadenitis followed by acute sialadenitis. Chronic sialadenitis can be confused with Warthin's tumor at cytology since both the lesions show a common lymphoid background. In the present study, the discordant case in this category was that of chronic sialadenitis which was finally confirmed as Warthins's tumor on histopathology. Failure to identify the dirty background and scant cellularity was the reason for the error in our case. Fersech et al.[6] in a study involving FNAC of 101 salivary gland lesions, three cases of Warthins's tumor were wrongly reported as chronic sialadenitis at FNAC. Acute sialadenitis was reported in 20/68 (15.2%) of cases in present study. All the cases underwent medical line of management and on follow-up of the majority of the lesions (90%) subsided. Only in two patients, in whom the swelling persisted, excision showed a consistent diagnosis. Sialadenosis was reported in 4/68 (3.05%) of cases. The smears of sialadenosis show acinar cells in cohesive, grape-like clusters with adipocytes and benign ductal cells.[11] Careful attention to the pattern and cellular details is required while diagnosing sialadenosis as it closely mimics acinic cell carcinoma.[11]

The benign category includes specific diagnostic entities such as pleomorphic adenoma, basal cell adenoma, and Warthins's tumor. DA of FNAC for benign lesions in the present study was high (100%), similar to that reported in other studies.

Pleomorphic adenoma is the most common benign tumor encountered in the salivary gland. In the present study, pleomorphic adenoma represented 78.57% of all benign tumors. The cytohistologic correlation was obtained in all cases, except in one case which showed pleomorphic adenoma with marked atypia and hence was placed in atypical category. The histopathology follow-up showed a low-grade mucoepidermoid carcinoma. Pitfalls occur in the diagnosis of pleomorphic adenoma due to the varied appearance of this tumor.[12],[13],[14] Anyone of the three components of pleomorphic adenoma, that is, ductal or myoepithelial cells, chondromyxoid fibrillary stroma may dominate the smears and cause errors. Hyaline globules are rarely seen in pleomorphic adenoma and are more commonly seen in adenoid cystic carcinoma, basal cell adenoma, and epithelial myoepithelial carcinoma. Attention should be paid to the relation between the cells and hyaline globules, melting of cells in hyaline lobules favor a pleomorphic adenoma, while sharp demarcation between the two favors adenoid cystic carcinoma. Similarly, hyaline globules in adenoid cystic carcinoma tend to be more hyaline and variable in size. The presence of characteristic chondromyxoid fibrillary stroma is highly characteristic and is a strong predictor of pleomorphic adenoma. Hence, a diligent and meticulous search of all the smears for the presence of chondromyxoid fibrillary stroma will avoid false-negative diagnosis.[15] Marked atypia at times can be seen in pleomorphic adenoma, which may lead to a misdiagnosis of carcinoma ex pleomorphic adenoma. Another well-recognized pitfall of FNAC is the difficulty to differentiate pleomorphic adenoma from low-grade mucoepidermoid carcinoma. The intermediate cells may be quite bland and the mucin may be confused with chondromyxoid fibrillary stroma. Goblet cells though, commonly seen in mucoepidermoid carcinoma, have been reported in pleomorphic adenoma. Furthermore, squamous metaplasia may occur secondary to infarct in a pleomorphic adenoma which may be worrisome.[16] However, plasmacytoid cells are more commonly seen in pleomorphic adenoma than in mucoepidermoid carcinoma. In a study by Kotwal et al.,[15] three out of four cases of mucoepidermoid carcinoma were misinterpreted as pleomorphic adenoma at FNAC. Multiple sampling from different sites is important to overcome the pitfall occurring owing to the variability in different parts of the same tumor.

There were two cases of basal cell adenoma in the present study. Both the cases showed concurrent diagnosis at histopathology. Aspirates of basal cell adenoma show small uniform basaloid cells with scanty cytoplasm and hyaline stroma. A tubular pattern and palisading can be appreciated and serve as a helpful clue.[17] Sometimes spindle morphology in basal cell adenoma raises a possibility of cellular pleomorphic adenoma or myoepithelioma.[17] Soomin et al.,[18] noted spindle cell morphology in 8 out of 14 basal cell adenoma. In all the cases with spindle cell morphology, a differential diagnosis of pleomorphic adenoma and myoepithelioma was rendered. Nevertheless, distinction among these is not mandatory owing to similar treatment approach for all benign tumors.[19] Adenoid cystic carcinoma, a close differential of basal cell adenoma, can be differentiated by the presence of cup-shaped structures, nuclear hyperchromasia, and nuclear membrane irregularity in the former lesion. In a study by Griffith et al.,[5] the malignancy rate in monomorphic basaloid neoplasm reached 35.5%. A total of six Warthin's tumor were diagnosed at FNAC in the present study. Out of these, four showed histopathology correlation. However, in remaining two cases, a final diagnosis of oncocytoma and oncocytic hyperplasia was made. Grifith et al.[5] referred to Warthin's tumor as “named entity” owing to a high DA for this tumor. Accurate diagnosis rest on finding of oncocytes and lymphocytes in a background of debritic material, absence or predominance of any of these may lead to erroneous diagnosis. Rarely Warthin's tumor can show squamous and sebaceous metaplasia. In both the discordant cases in the present study, plenty of bare nuclei in the background were mistaken for lymphocytes. However, we failed to notice the absence of debritic material in the background. Diagnosis of oncocytic lesions at cytology is a challenge due to focal sampling as oncocytic change is seen in a variety of salivary gland lesions.[19]

None of the cases reported in the benign category were malignant on surgical follow-up. Discordant diagnosis in this category was related to subtyping error and did not alter the management, ROM was 0%. Griffith et al.[5] and Rohilla et al.[10] reported a ROM of 2% and 7.3%, respectively.

Atypical cells of undetermined significance are a heterogeneous group in which most of the cases are reactive, however, the cellular details are masked by blood or drying artifact. Cases, in which definitive distinction between reactive and neoplastic process is not possible, are included in this category. One of the major challenges of FNAC of salivary gland lesions is the ability to distinguish benign mucinous cyst from low-grade mucoepidermoid carcinoma. The cystic lesions pose a nightmare to cytopathologists. Apart from benign mucinous cyst, many other tumors may show cystic change such as mucoepidermoid carcinoma, acinic cell carcinoma, and adenoid cystic carcinoma. Other lesions that can cause FN diagnosis of mucoepidermoid carcinoma are Warthin's tumor and pleomorphic adenoma. The ROM in this category in the present study was 100%. Similar findings were reported by Rohilla et al.[10] who classified one case in this category. Rossi et al.[9] reported 15 cases in atypical category with a ROM of 53%. This category essentially can come as a savior for the cytopathologist in difficult situations, at the same time, it alerts the clinician of the possible risk and hence deals with more caution in these cases. The ROM for this category has not been defined accurately, nevertheless further large-scale studies and literature review is required for delineating this entity and determining the true ROM.

The current study had two cases of salivary neoplasm of uncertain malignant potential (SNUMP), one was a basal cell adenoma in which high cellularity was a cause of concern, hence it was reported as basaloid neoplasm which turned out to be basal cell adenoma at histopathology. Another case was reported as pleomorphic adenoma with atypia. However, a discordant diagnosis of low-grade mucoepidermoid carcinoma was obtained at histopathology. Several authors have enunciated the difficulty in distinguishing pleomorphic adenoma from mucoepidermoid carcinoma. In the present study, mucin was confused with CMF stroma. Rossi et al.,[9] in their study, reported eight cases of pleomorphic adenoma at FNAC, which were diagnosed as malignant on surgical follow-up leading to a high false-positive rate of 17.4%. Other malignant lesions that can be misinterpreted as pleomorphic adenoma at FNAC are adenoid cystic carcinoma and epithelial myoepithelial carcinoma. The ROM for this category range from 2% to 50%. In the present study, the ROM was 50% similar to that reported by Rohilla et al.[10]

The suspicious category in the present study included three cases, all of which showed a cytohistologic correlation. The ROM was 100%, higher than that reported in literature, that is, 60% and 79% by MSRSGC and Rossi et al.,[9] respectively. All three cases of low-grade mucoepidermoid carcinoma showed obvious features of malignancy but the lack of significant atypia prompted us to classify these lesions in suspicious rather than malignant category.

The overall prevalence of malignancy in present study was 17.5%, well within the range observed by other authors (15–32%).[17],[18],[19] In the present study, mucoepidermoid carcinoma was the most common malignancy accounting for 52% of all malignant lesions followed by acinic cell carcinoma (13.04%). Many studies report similar incidence rates.[20] The cytohistologic correlation for this category was 100%. ROM was 100% which was similar to that proposed by MSRSGC and reported by Rossi et al.[9] and Griffith et al.[5] Nagel et al.[21] retrospectively reviewed 58 cases of acinic cell carcinoma diagnosed at FNAC. Apart from acinar cell, oncocytic cells, nonspecific goblet cells, plenty of bare nuclei and lymphocytes are seen in 10% of cases are all clue to the diagnosis of acinic cell carcinoma.[21] DA of FNAC for acinic cell carcinoma has been reported to be 88%. Deviant subtypes may at times pose diagnostic difficulties. Errors occur mainly due to the inability to identify the cellularity and architecture.

ROM for various categories in the present study shows similarities with the study of Rohilla et al.[10] but deviates from that proposed by MSRSGC especially the atypical and SNUMP category. Atypical category, in the present study, included only one case, which showed scanty cellularity, on review, comprising intermediate cells and should have been placed in the SNUMP category. The ROM in atypical category is still a matter of debate and should range between 10 and 35%. Rohilla et al.[10] reported ROM of 50% in atypical category, which was similar to that in the present study. The major challenges of calculating the ROM in various categories and its clinical applicability are the limited number of surgical follow-up obtained. Most of the time only neoplastic lesions are excised leading to overestimation of ROM. This was a bias noted in a study by Rohilla et al.[10] However, in the current study, surgical follow-up was obtained in 48% of the cases out of which 26% were nonneoplastic.

The false-positive rate in the present study was 0%. Various studies have reported a false positive rate ranging from 0% to 10%.[9] FN rate in the present study was 1.52% lower than that reported in the literature (2.2–24.5%).[22]

As seen in [Table 2], we can see that the diagnostic accuracy for all the categories was good except for atypical and SNUMP categories. This way we attribute to the ambiguity over the specific criteria, inherent nature of the lesions, and the technical issues. This being said the extremely low DA in atypical category in the present study is attributed to the fact that only one case was classified as atypical which was FN leading to DA of 0%.

The diagnostic efficacy of the Milan system in present study was similar to that reported in the literature.[23],[24],[25],[26],[27] The comparison of ROM for individual category of Milan with various studies reported in literature is shown in [Table 3].[23],[24],[25],[26],[27],[28],[29]

The potential limitation of the study was the small sample size in a limited time frame. The high efficacy of FNAC obtained in the present study, when MSRSGC was applied, confirms the usefulness of this scheme in reporting salivary gland lesions. A risk-based stratification is essential in the present era to guide and alert the clinician about the subsequent management plan and the ROM.

Informed consent



Authors acknowledge the support provided by Dr MG Shivaramu, Dean, Health Sciences (Medical), Adichunchanagiri Institute of Medical Sciences, Adichunchanagiri University.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Orell SR, Sterrett GF, Whitaker D, Klijanienko J. Head and neck; salivary glands. In: Orell SR, Sterrett GF, Whitaker D, editors. Fine Needle Aspiration Cytology. 4th ed. Edinburgh: Churchill Livingstone-Elsevier; 2005. p. 53-69.
2Rajwanshi A, Gupta K, Gupta N, Shukla R, Srinivasan R, Nijhawan R, et al. Fine-needle aspiration cytology of salivary glands: Diagnostic pitfalls-revisited. Diagn Cytopathol 2006;34:580-4.
3Tyagi R. Dey P. Diagnostic problems of salivary gland tumors. Diagn Cytopathol 2015;43:495-509.
4Tahoun N, Ezzat N. Diagnostic accuracy and pitfalls of preoperative fine needle aspiration cytology in salivary gland lesions. J Egypt Natl Canc Inst 2008;20:358-8.
5Griffith CC, Schmitt AC, Pantanowitz L, Monaco SE. A pattern-based risk-stratification scheme for salivary gland cytology: A multi-institutional, interobserver variability study to determine applicability. Cancer Cytopathol 2017;125:776-5
6Amita K, Vijay Shankar S, Sanjay M, Sarvesh BM. Effectiveness of the pattern-based approach in the cytodiagnosis of salivary gland lesions. Acta Cytol 2016;60:107-7.
77 Faquin WC, Powers C. Salivary gland cytopathology. In: Algorithmic Approach to Salivary Gland FNA: An Overview. New York: Springer Science & Business Media; 2008. p. 29-40.
8Griffith CC, Pai RK, Schneider F, Duvvuri U, Ferris RL, Johnson JT, et al. Salivary gland tumor fine-needle aspiration cytology: A proposal for a risk stratification classification. Am J Clin Pathol 2015;143:839-53.
9Rossi ED, Wong LQ, Bizzarro T, Petrone G, Mule A, Fadda G, et al. The impact of FNAC in the management of salivary gland lesions: Institutional experiences leading to a risk-based classification scheme. Cancer Cytopathol 2016;124:388-96.
10Rohilla M, Singh P, Rajwanshi A, Gupta N, Srinivasan R, Dey P, et al. Three-year cytohistological correlation of salivary gland FNA cytology at a tertiary center with the application of the Milan system for risk stratification. Cancer Cytopathol 2017;125:767-75.
11Ascoli V, Albedi FM, Blasiis RD, Nardi F. Sialadenosis of the parotid gland: Report of four cases diagnosed by fine-needle aspiration cytology. Diagn Cytopathol 1993:9:151-5.
12Ryan RE Jr, DeSanto LW, Weiland LH, Devine KD, Beahrs OH. Cellular mixed tumors of the salivary glands. Arch Otolaryngol 1978;104:451-3.
13Verma K, Kapila K. Role of fine needle aspiration cytology in diagnosis of pleomorphic adenomas. Cytopathology 2002;13:121-7.
14Rajwanshi A, Gupta K, Gupta N, Shukla R, Srinivasan R, Nijhawan R, et al. Fine-needle aspiration cytology of salivary glands: Diagnostic pitfalls-Revisited. Diagn Cytopathol 2006;34:580-4.
15Kotwal M, Gaikwad S, Patil R, Munshi M, Bobhate S. FNAC of salivary gland-A useful tool in preoperative diagnosis or a cytopathogist's riddle?. J Cytol 2007;24:85-8.
16Goulart MC, Freitas-Faria P, Goulart GR, Oliveira AM, Carlos-Bregni R, Soares CT. Pleomorphic adenoma with extensive squamous metaplasia and keratin cyst formations in minor salivary gland: A case report. J Appl Oral Sci 2011;19:182-8.
17Hara H, Oyama T, Saku T. Fine needle aspiration cytology of basal cell adenoma of the salivary gland. Acta Cytol 2007;51:685-91.
18. Ahn S, Kim Y, Oh L. Fine Needle aspiration cytology of benign salivary gland tumors with myoepithelial cell participation: An institutional experience of 575 cases. Acta Cytologica 2013;57:567-74.
19Chakrabarti I, Basu A, Ghosh N. Oncocytic lesion of parotid gland: A dilemma for cytopathologists. J Cytol 2012;29:80-2.
20Vasudevan G, Bishnu A, Singh BMK, SinghVK. Mucoepidermoid carcinoma of salivary gland: Limitations and pitfalls on FNA. J Clin Diagn Res 2017;11:ER04-06.
21Nagel H, Laskawi R, Büter JJ, Schröder M, Chilla R, Droese M. Cytologic diagnosis of acinic-cell carcinoma of salivary glands. Diagn Cytopathol 1997;16:402-12.
22Aisagbonhi OA, Tulecke MA, Wilbur DC, AtoussaGoldar-Najafi W, Iqbal S, Sadow PM, et al. Fine-needle aspiration of epithelial-myoepithelial carcinoma of the parotid gland with prominent adenoid cystic carcinoma–like cribriform features: Avoiding a diagnostic pitfall. Am J Clin Pathol 2016;146:741-6.
23Vallonthaiel AG, Kaushal S, Jangir H, Rajendran HK. Application of the Milan system for risk stratification and its comparison with a previous reporting system of parotid gland cytopathology in a tertiary care centre. Acta Cytol 2018;17:1-8.
24Thiryayi SA, Low YX, Shelton D, Narine N, Slater D, Rana DN. A retrospective 3-year study of salivary gland FNAC with categorisation using the Milan reporting system. Cytopathology 2018;29:343-8.
25Viswanathan K, Sung S, Scognamiglio T, Yang GCH, Siddiqui MT, Rao RA. The role of the Milan system for reporting salivary gland cytopathology: A 5-year institutional experience. Cancer Cytopathol 2018;126:541-51.
26Pujani M, Chauhan V, Agarwal C, Raychaudhuri S, Singh K. A critical appraisal of the Milan system for reporting salivary gland cytology (MSRSGC) with histological correlation over a 3-year period: Indian scenario. Diagn Cytopathol 2018;47:382-8.
27Farahani SJ, Baloch Z. Retrospective assessment of the effectiveness of the Milan system for reporting salivary glandcytology: A systematic review and meta-analysis of published literature. Diagn Cytopathol 2019;47:67-87.
28Layfield LJ, Baloch ZW, Hirschowitz SL, Rossi ED. Impact on clinical follow-up of the Milan System for salivary gland cytology: A comparison with a traditional diagnostic classification. Cytopathology 2018;29:335-42.
29Hollyfield JM, O'Connor SM, Maygarden SJ, Greene KG, Scanga LR, Tang S, et al. Northern Italy in the American South: Assessing interobserver reliability within the Milan System for Reporting SalivaryGland Cytopathology. Cancer Cytopathol 2018;126:390-6.