Moving Toward a Systematic Approach for Reporting Salivary Gland Cytopathology: An Institutional Experience and Literature Review.
In the daily cytopathology practice at Temple University Hospital (Philadelphia, Pennsylvania), we had long recognized the problem of inconsistent diagnostic terminology. Together with our head and neck surgeons, we developed a homegrown reporting system and adopted it into our practice with good results. Our system is quite similar to the proposed Milan System for Reporting Salivary Gland Cytopathology. The present study aimed to summarize our institutional experience with the institutional 6-tier salivary gland cytopathology reporting system, as an initiative to promote collaborative work toward a consensus on a reporting system that is succinct, unambiguous, and clinically helpful. We believe our experience will be of value to the adoption of the Milan System for Reporting Salivary Gland Cytopathology.
Slide Review of Institutional Salivary Gland FNA Cases
From the electronic files of Temple University Hospital's pathology department, 154 consecutive cases of salivary gland FNA were identified. Cases with insufficient diagnostic material (broken or missing slides, slides with uninterpretable cells; 47 cases) were excluded from further analysis. Slides of the remaining 107 cases (during a period of 15 years, from 1999 to 2014) from this list were retrieved and reviewed. After documentation and confirmation of original diagnoses, the FNAs were reclassified into the 6-tier classification system.
This study complied with the Declaration of Helsinki and received approval from the Institutional Review Board.
A 6-Tier Classification System
A 6-tier classification system was proposed and evaluated in the present study based on our experience and review of the 107 salivary gland FNAs from Temple University Hospital. The details of the 6-tier classification system (4,9) are listed in Table 1. In brief, the classification system evaluated in the present study categorized salivary gland cytopathology into the following diagnostic groups: unsatisfactory, negative for neoplasm, lesion of unknown significance, benign neoplasm, suspicious for malignancy, and positive for malignancy.
Correlation of Reclassified Salivary Gland FNA and Surgical Pathology Diagnoses
Subsequent excision specimens were available for 57 of the 107 FNA cases reviewed, and these histology slides were retrieved from our archive. The FNA specimens were then de-identified, reviewed, and categorized by 2 experienced cytopathologists individually using the 6-tier classification system. Interobserver variability was assessed with the [kappa] statistic. Excision specimen slides were reviewed by an experienced head and neck surgical pathologist and compared with reconciled cytologic diagnoses for cytologyhistology correlation. This subset of cases also allowed us to calculate risk of malignancy (ROM) and risk of neoplasm (RON) for different diagnostic categories in the classification system.
Reclassification of Salivary Gland FNA Using the 6-Tier Classification System
The original diagnoses of the 107 salivary gland FNAs included 15 different diagnostic designations ranging from blood only to abnormal/insufficient to positive for malignancy. After reclassification, our cohort included 8 unsatisfactory cases, 7 negative for neoplasm cases, 19 lesions of unknown significance, 50 benign neoplasm cases, 6 suspicious for malignancy, and 17 malignant neoplasm cases (Figure). The mean age of our cohort was 59.3 years (range, 21-85 years); the male to female ratio was 41:66. Table 2 lists the final histologic diagnoses of all the cases for which subsequent excision specimens were available by category.
ROM and RON of the Reclassified Salivary Gland FNAs
As shown in Table 3, of the 13 negative for neoplasm FNAs, 10 (77%) were confirmed to be negative for neoplasm on the excision specimens, and the remaining 3 (23%) turned out to be benign neoplasms. The ROM and RON for this category were 0% and 23%, respectively. Twenty-seven of 29 FNA cases (93%) identified as benign neoplasm were confirmed after excision as well, and the remaining 2 (7%) were diagnosed as malignancy on the excision specimens. The ROM and RON for this category were 7% and 100%, respectively. Of the 3 cases identified as suspicious for malignancy, 1 was confirmed to be pleomorphic adenoma (PA), and the other 2 were mucoepidermoid carcinomas. The ROM and RON for this category were 67% and 100%, respectively. Ten of the 12 cases (83%) categorized as positive for malignancy by cytology were confirmed to be malignancies after excision, whereas the remaining 2 cases (17%) turned out to be negative for neoplasm. The ROM and RON for this category were 83% and 83%, respectively. No cases categorized as unsatisfactory or lesion of unknown significance on FNA had histologic follow-up.
Performance of the 6-Tier Classification System in Diagnosing Malignancies
Table 4 is a 2 X 2 contingency table showing the agreement between FNA and histology in terms of diagnosing salivary gland malignancies. The 6-tier cytology classification system performed well, with a sensitivity of 83%, a specificity of 93%, a positive predictive value of 80%, and a negative predictive value of 95%.
Table 3 is a comparison of reconciled FNA and histology diagnoses. In general, FNA diagnoses made based on the proposed 6-tier classification system were consistent with histology diagnoses made on the excision specimens.
Interobserver Variability With the 6-Tier Classification System
Interobserver variability was assessed by comparing the agreement between the 2 cytopathologists on categorizing the 57 cases with both FNA and excision specimens using the 6-tier classification system. The results are summarized in Table 5. The numbers in the cells along the diagonal axis from upper left to lower right represent numbers of cases (by categories proposed in the 6-tier classification system) for which both cytopathologists agreed on the diagnosis. The overall agreement was 53 of 57, or 93%. The k statistic was 0.893, with an SE of 0.051 and a 95% CI of 0.794-0.992, indicating substantial agreement between the 2 cytopathologists, who disagreed on the salivary FNA diagnosis in only 4 cases (Table 6).
DISCUSSION AND LITERATURE REVIEW
The proposed 6-tier classification system for salivary gland cytology is based on the concept of reporting cytopathology on a discrete 5-step scale of increasing likelihood/certainty of malignancy, with an additional category for inadequate/ nondiagnostic specimens. It provided a succinct and ROM-based tool with which substantial interobserver agreement (Table 5) between the cytopathologists in reporting salivary gland cytologic findings could be achieved. It is worth noting that interobserver agreement was performed between only 2 cytopathologists. This is a limitation to the study because having additional cytopathologists would most likely cause a decline in the k statistic. The ROM increased from 0% (0 of 13) to 7% (2 of 29), 67% (2 of 3), and 83% (10 of 12) for categories of negative for neoplasm, benign neoplasm, suspicious for malignancy, and positive for malignancy. In general, FNA diagnoses made based on the proposed 6-tier classification system were consistent with histology diagnoses for the disease spectrum reported in the present study. The sensitivity, specificity, positive predictive value, and negative predictive value of diagnosing malignancies with the proposed classification system were 86% (12 of 14), 93% (40 of 43), 80% (12 of 15), and 95% (40 of 42), respectively.
Several cytology reporting systems have been proposed and used in practice, including the current Bethesda System for Reporting Thyroid Cytopathology, (10) the Paris System for Reporting Urinary Cytology, (11) a recently proposed system for reporting endometrial cytology, (12) and the current Bethesda System for Reporting Cervical Cytology. (13) Although these reporting systems were originally designed to report cytopathology of different organs, the diagnostic categories defined in these systems can all be fit roughly into 6 tiers similar to those proposed in the present study based on the likelihood/certainty of malignancy (Table 7). The fact that all these individual reporting systems fit into a 6-tier framework may not just be a coincidence, but may be due to the nature of cytopathology as a diagnostic modality as well as influence from the original Papanicolaou classification for vaginal smear-based cytologic diagnosis of uterine cancer. (14)
A different 6-tier classification system was proposed for parotid gland cytology in 2014 and was later modified into a more sophisticated 9-tier system to report salivary gland cytopathology. (15,16) The 9 categories are inadequate (Sal 1, previous P1), cyst contents only (Sal 1c, previous P1), inflammatory nonneoplastic (Sal 2I, previous P2), benign neoplastic (Sal 2N, previous P3), atypical (Sal 3, previous P4), suspicious (Sal 4, previous P4), primary salivary gland malignancy (Sal 5P, previous P5), malignancy not otherwise specified (Sal 5NOS, previous P5), and metastatic malignancy (Sal 5M, previous P6). The 9-tier system has an overall sensitivity of 61%, specificity of 97%, positive predictive value of 82%, and negative predictive value of 91% for the diagnosis of malignancy. A systematic review and meta-analysis (17) on the diagnostic accuracy of FNA for parotid gland malignancies reported a sensitivity of 80% and a specificity of 97%. The 9-tier system falls short on sensitivity. Our proposed categorization system, however, demonstrates a sensitivity of 83% and a specificity of 93%, which are on par with the meta-analysis. Several additional differences have been noted between the present and the 9-tier system studies. First, a substantial interobserver agreement was achieved by using our proposed categorization system; because only the cytology reports were reviewed in the 9-tier system study, interobserver agreement could therefore not be assessed. Second, histology on the excision specimen served as the final diagnosis in the present study to evaluate the performance of our proposed categorization system; in the 9-tier system study, however, approximately 20% of the enrolled cases lacked histology diagnoses on excision specimens, and clinical diagnoses were used instead, which could have interfered with the performance of the categorization system.
Our proposed 6-tier system shares many similarities with the Milan System for Reporting Salivary Gland Cytopathology, (7,8) which also has 6 tiers and includes the categories of nondiagnostic, nonneoplastic, benign neoplasm, salivary gland neoplasm of undetermined malignant potential, suspicious for malignant neoplasm, and malignant. In a recent meta-analysis using the Milan system, (18) the ROMs for nonneoplastic, benign neoplasm, suspicious for malignant neoplasm, and malignant were 10% [+ or -] 6%, 3% [+ or -] 1%, 59% [+ or -] 20%, and 92% [+ or -] 4%, respectively. Although the size of our study group was relatively small, which was a major limitation, our results by category were consistent with those from the meta-analysis.
The definition of adequacy is a common element to all systems (the 9-tier system, the Milan System for Reporting Salivary Gland Cytopathology, and our proposed 6-tier system) that requires further investigation. The 9-tier system defines a cytology specimen as inadequate if the specimen is nondiagnostic, paucicellular, cyst aspirate only, blood only, necrotic debris/pus only, normal salivary gland tissue (technical miss or sialadenosis), or preparation artifact. (16) The definition of inadequacy in our proposed system is derived anecdotally from our practice and is more quantitative, and it can be applied more objectively in daily practice. It is without doubt, however, that a consensus on the definition of adequacy is necessary among cytopathologists in the field before a prospective study comparing various reporting systems can be conducted. We believe that with widespread use of the Milan System for Reporting Salivary Gland Cytopathology worldwide, consensus on issues like adequacy will be reached in the near future.
As an initiative to promote collaborative work toward a consensus on a reporting system for salivary gland cytology, the proposed 6-tier classification system provides a succinct and ROM-based tool with satisfactory interobserver agreement in reporting salivary gland cytology.
(1.) Seethala RR, LiVolsi VA, Baloch ZW. Relative accuracy of fine-needle aspiration and frozen section in the diagnosis of lesions of the parotid gland. Head Neck. 2005;27(3):217-223.
(2.) Zbaren P, Guelat D, Loosli H, Stauffer E. Parotid tumors: fine-needle aspiration and/or frozen section. Otolaryngol Head NeckSurg. 2008;139(6):811-815.
(3.) Carrillo JF, Ramirez R, Flores L, et al. Diagnostic accuracy of fine needle aspiration biopsy in preoperative diagnosis of patients with parotid gland masses. J Surg Oncol. 2009;100(2):133-138.
(4.) Wang H, Fundakowski C, Khurana JS, Jhala N. Fine-needle aspiration biopsy of salivary gland lesions. Arch Pathol Lab Med. 2015;139(12):1491-1497.
(5.) Liu S, Parajuli S, Hotchandani N, et al. Fine needle aspiration diagnosis of non-epithelial lesions of the major salivary glands. Int J Clin Exp Pathol. 2016;9(7):7164-7171.
(6.) El-Naggar AK, Chan JKC, Grandis JR, Takata T, Slooweg PJ. WHO Classification of Head and Neck Tumours. Lyon, France: IARC Press; 2017.
(7.) Baloch ZW, Faquin WC, Layfield LJ. Is it time to develop a tiered classification scheme for salivary gland fine-needle aspiration specimens? Diagn Cytopathol. 2017;45(4):285-286.
(8.) Rossi ED, Faquin WC, Baloch Z, et al. The Milan System for Reporting Salivary Gland Cytopathology: analysis and suggestions of initial survey. Cancer. 2017;125(10):757-766.
(9.) Fundakowski C, Castano J, Abouyared M, et al. The role of indeterminate fine-needle biopsy in the diagnosis of parotid malignancy. Laryngoscope. 2014; 124(3):678-681.
(10.) Baloch ZW, Alexander EK, Gharib H, Raab SS. Overview of diagnostic terminology and reporting. In: Ali SZ, Cibas ES, eds. The Bethesda System for Reporting Thyroid Cytopathology. New York, NY: Springer; 2010:1-4.
(11.) Rosenthal DL, Wojcik EM, Kurtycz DFI, eds. The Paris System for Reporting Urinary Cytology. Cham, Switzerland: Springer; 2016.
(12.) Margari N, Pouliakis A, Anoinos D, et al. A reporting system for endometrial cytology: cytomorphologic criteria--implied risk of malignancy. Diagn Cytopathol. 2016;44(11):888-901.
(13.) Nayar R, Wilbur DC, eds. The Bethesda System for Reporting Cervical Cytology. 3rd ed. Cham, Switzerland: Springer; 2015.
(14.) Papanicolaou GN, Trunt HF. Diagnosis of Uterine Cancer by the Vaginal Smear. New York, NY: Commonwealth Fund; 1943.
(15.) Bajwa MS, Mitchell DA, Brennan PA. Is it time we adopted a classification for parotid gland cytology? Br J Oral Maxillofac Surg. 2014;52(2):99-101.
(16.) Bajwa MS, Rose SJ, Mairembam P, et al. Feasibility of a novel classification for parotid gland cytology: a retrospective review of 512 cytology reports taken from 4 United Kingdom general hospitals. Head Neck. 2016;38(11):1596-1603.
(17.) Schmidt RL, Hunt JP, Hall BJ, Wilson AR, Layfield LJ. A systematic review and meta-analysis of the diagnostic accuracy of frozen section for parotid gland lesions. Am J Clin Pathol. 2011;136(5):729-738.
(18.) Wei S, Layfield LJ, LiVolsi VA, Montone KT, Baloch ZW. Reporting of fine needle aspiration (FNA) specimens of salivary gland lesions: a comprehensive review. Diagn Cytopathol. 2017;45(9):820-827.
Xunda Luo, MD, PhD; Nirag Jhala, MD; Jasvir S. Khurana, MD; Christopher Fundakowski, MD; Darshana N. Jhala, MD; He Wang, MD, PhD
Accepted for publication February 8, 2018.
Published online July 25, 2018.
From the Departments of Pathology (Drs Luo, N. Jhala, and Khurana) and Otolaryngology (Dr Fundakowski), Temple University School of Medicine, Philadelphia, Pennsylvania; the Department of Pathology and Laboratory Medicine, University of Pennsylvania Perelman School of Medicine, Philadelphia (Dr D. N. Jhala); and the Department of Pathology and Laboratory Medicine, Robert Wood Johnson Medical School, Rutgers University, New Brunswick, New Jersey (Dr Wang).
The authors have no relevant financial interest in the products or companies described in this article.
Presented in part at the First Princeton Integrated Pathology Symposium; February 9, 2014; Princeton, New Jersey.
Corresponding author: He Wang, MD, PhD, Department of Pathology, Robert Wood Johnson Medical School, Rutgers University, 125 Paterson St, New Brunswick, NJ 08901 (email: hw423@ rwjms.rutgers.edu).
Caption: The role of indeterminate fine-needle aspiration (FNA) in the diagnosis of parotid malignancy. A, Original FNA diagnoses. B, Revised FNA diagnoses. Abbreviations: c/w, consistent with; LOUS, lesion of unknown significance; Neg, negative; PA, pleomorphic adenoma; Pos, positive; Susp, suspicious; Unsat, unsatisfactory; WT, Warthin tumor.
Table 1. Proposed 6-Tier Classification System for Salivary Gland Cytopathology I. Unsatisfactory 1. Fewer than 4 clusters (at least 10 cells per cluster) of epithelial and/or mesenchymal cells 2. Fewer than 200 cells if only hematopoietic cells are present in the specimen 3. Specimen with only fluid, keratin debris, or other extracellular deposits, and no additional radiologic and clinical information available II. Negative for neoplasm 1. No neoplastic cells noted III. Lesion of unknown significance 1. Lesions with clear cellular atypia 2. Reactive versus neoplastic processes cannot be distinguished despite all differential efforts IV. Benign neoplasm 1. Combined cytologic features of epithelial cells, myoepithelial cells, mesenchymal cells, stroma, and smear background indicate unequivocally neoplastic process 2. Cytologic features including but not limited to high cellularity, crowding within the same cluster, nuclear pleomorphism, hyperchromasia, and irregular nuclear contour V. Suspicious for malignancy 1. Presence of features suggestive of but not sufficient to make a conclusive diagnosis for malignancy despite all differential efforts including ancillary molecular and immunohistochemical testing VI. Positive for malignancy 1. Combined cytologic features in epithelial cells, myoepithelial cells, and/or mesenchymal cells clearly indicate malignancy 2. Cytologic features including but not limited to crowding and disorientation of cells in the same cluster, significantly increased nuclear to cytoplasmic ratio, uneven distribution of nuclear chromatin, irregular nuclear contour, nuclear pleomorphism, prominent nucleoli, and necrotic background in the smear Table 2. Histology Diagnosis by Category Histology Diagnosis No. Details No. Nonneoplastic 11 Lymphoepithelial cyst 3 Inflammation 4 Epidermal inclusion cyst 1 Granuloma 1 Ectopic thyroid 1 Benign salivary gland 1 Benign 32 Monomorphic adenoma 1 neoplasm Pleomorphic adenoma 16 Warthin tumor 10 Granular cell tumor 2 Lipoma(tosis)/hibernoma 2 Oncocytoma 1 Malignant 14 Adenocarcinoma, NOS 4 neoplasm Mucoepidermoid carcinoma 4 Diffuse large B cell lymphoma 1 Carcinoma ex pleomorphic adenoma 1 Squamous cell carcinoma 2 Poorly differentiated carcinoma 1 Metastatic carcinoma 1 Abbreviation: NOS, not otherwise specified. Table 3. Cytology-Histology Correlation Reconciliated FNA Diagnostic No. Histology Diagnosis Category I. Unsatisfactory 0 Unsatisfactory II. Negative for neoplasm 13 Negative for neoplasm Benign neoplasm III. Lesion of unknown 0 Lesion of unknown significance significance IV. Benign neoplasm 29 Benign neoplasm Positive for malignancy V. Suspicious for malignancy 3 Benign neoplasm Positive for malignancy VI. Positive for malignancy 12 Negative for neoplasm Benign neoplasm Positive for malignancy Reconciliated FNA Diagnostic No. Comments Category I. Unsatisfactory 0 II. Negative for neoplasm 10 3 Warthin tumor (2), hibernoma (1) III. Lesion of unknown 0 significance IV. Benign neoplasm 27 2 Mucoepidermoid carcinoma (1), carcinoma ex pleomorphic adenoma (1) V. Suspicious for malignancy 1 Pleomorphic adenoma 2 Mucoepidermoid carcinoma VI. Positive for malignancy 1 Granuloma 1 Lipomatosis 10 Abbreviation: FNA, fine-needle aspiration. Table 4. Performance of the Classification System Histology FNA Benign Malignant Analysis I-IV 40 2 NPV = 95.2% V, VI 3 12 PPV = 80% Spe = 93.0% Sen = 83.3% Abbreviations: FNA, fine-needle aspiration;NPV, negative predictive value;PPV, positive predictive value;Sen, sensitivity;Spe, specificity. Table 5. Interobserver Variability Using the 6-Tier Classification System Cytopathologist B Cytopathologist A I II III IV V VI Total I 0 0 0 0 0 0 0 II 0 12 0 1 0 0 13 III 0 0 0 2 0 0 2 IV 0 0 0 27 0 0 27 V 0 0 0 0 3 1 4 VI 0 0 0 0 0 11 11 Total 0 12 0 30 3 12 57 Table 6. Cases With Discrepancies Between the 2 Cytopathologists Age, Classification Patient y/Sex CDx1 1 P13 40/F Negative for neoplasm II P14 63/F Lesion of unknown significance III P15 65/F Lesion of unknown significance III P46 85/F Suspicious for malignancy V Classification Patient CDx2 2 P13 Benign neoplasm IV P14 Benign neoplasm IV P15 Benign neoplasm IV P46 Positive for malignancy VI Patient HDx P13 Epidermal inclusion cyst P14 Granular cell tumor P15 Granular cell tumor P46 Low-grade malignant myoepithelioma Abbreviations: CDx, cytology diagnosis;HDx, histology diagnosis. Table 7. Fit of Current Cytology Reporting Systems Into a 6-Tier Framework Proposed System in the Present Study Bethesda System for Salivary Gland for Thyroid Category (a) Cytology Reporting Cytology Reporting I Unsatisfactory Nondiagnostic or unsatisfactory II Negative for Benign neoplasm III Lesion of unknown Atypical of undetermined significance significance or follicular lesion of undetermined significance IV Benign neoplasm Follicular neoplasm or suspicious for a follicular neoplasm V Suspicious for Suspicious for malignancy malignancy VI Positive for Malignant malignancy Paris System System for for Urinary Endometrial Category (a) Cytology Reporting Cytology Reporting I Inadequate Nondiagnostic or unsatisfactory II Negative for high-grade Without evidence urothelial carcinoma of hyperplasia or malignancy III Atypical urothelial cells Atypical cells of endometrium of undetermined significance IV Low-grade urothelial Atypical cells of carcinoma low probability for malignancy V Suspicious for high-grade Atypical cells of urothelial carcinoma high probability for malignancy VI High-grade urothelial Malignant carcinoma, and other malignancies primary and metastatic Bethesda System for Cervical Category (a) Cytology Reporting I Inadequate II Negative for intraepithelial lesion or malignancy III ASC-US, atypical endocervical cells, atypical endometrial cells, atypical glandular cells IV LSIL, atypical endocervical cells favor neoplastic, atypical glandular cells favor neoplastic V ASC-H, HSIL, endocervical adenocarcinoma in situ VI Squamous cell carcinoma, adenocarcinoma, other malignant neoplasms Abbreviations: ASC-H, atypical squamous cells, cannot rule out high-grade squamous intraepithelial lesion;ASC-US, atypical squamous cells of undetermined significance; HSIL, high-grade squamous intraepithelial lesion;LSIL, low-grade squamous intraepithelial lesion. (a) Categories are numbered in order of increasing likelihood/ certainty of malignancy.
|Printer friendly Cite/link Email Feedback|
|Author:||Luo, Xunda; Jhala, Nirag; Khurana, Jasvir S.; Fundakowski, Christopher; Jhala, Darshana N.; Wang, He|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Date:||Jun 1, 2019|
|Previous Article:||The Evolving Field of Cytopathology and Its Expanding Role in Pathologic Practice.|
|Next Article:||Molecular Testing of Non-Small Cell Lung Carcinoma Diagnosed by Endobronchial Ultrasound-Guided Transbronchial Fine-Needle Aspiration: The Cleveland...|