A Study of Reproducibility of Papanicolaou Society of Cytopathology System of Reporting Respiratory Cytology-A Single Institutional Experience on Image Guided Aspiration Cytology


 Background: Cytologic examination of specimens obtained from the respiratory tract is a primary and frequently the initial diagnostic technique performed in patients with pulmonary abnormalities. Fine needle aspiration (FNA) is extensively used for diagnosis of pulmonary lesions. The Papanicolaou Society of Cytopathology (PSC) issued a new classification for respiratory cytology and criteria, risk of malignancy, post-cytologic diagnosis management and follow-up. Methods: Respiratory FNA specimens obtained between January 2015 to March 2021 were reviewed and reclassified according to PSC guidelines. Cytologic category as per PSC system was assigned after reviewing by two pathologists. Risk of malignancy for each category was calculated for cases where biopsy was available for correlation. Results: Three eighty-four samples were classified as non-diagnostic (30.5%), negative for malignancy (13.3%), atypical (0.5%), neoplastic (benign/ low malignant potential) (0%), suspicious for malignancy (5.7%%) and malignancy (50%). Risk of malignancy for malignant category (Category V and Category VI) was 64.1% and for non-malignant (Category I to Category IV) cases was 35.9%. Overall, there was lesser incidence of atypical category and more specific diagnoses were possible on cytology in non-malignant category. Sensitivity and specificity for malignant cases was 99.33% and 100% respectively. Conclusion: Our study substantiates that use of PSC guidelines improves the overall reporting of respiratory cytology due to the use of standardized terminology.

biopsy was available for correlation.
Results: Three eighty-four samples were classi ed as non-diagnostic (30.5%), negative for malignancy (13.3%), atypical (0.5%), neoplastic (benign/ low malignant potential) (0%), suspicious for malignancy (5.7%%) and malignancy (50%). Risk of malignancy for malignant category (Category V and Category VI) was 64.1% and for non-malignant (Category I to Category IV) cases was 35.9%. Overall, there was lesser incidence of atypical category and more speci c diagnoses were possible on cytology in non-malignant category. Sensitivity and speci city for malignant cases was 99.33% and 100% respectively.
Conclusion: Our study substantiates that use of PSC guidelines improves the overall reporting of respiratory cytology due to the use of standardized terminology.

Background
Worldwide, Lung cancer is one of the leading causes of incidence as well as death. 1 A tactful tool in the diagnosis of lung malignancies is cytology. 2 Cytological examination of specimens obtained from the respiratory tract is a primary and frequently the initial diagnostic technique performed in patients with pulmonary abnormalities.
FNAC is commonly considered the rst choice for diagnosis of lesions located in the mediastinum, deep hilar lesions, pulmonary apex, medial upper lobe or peripheral lung, from where a sample can be attained using ne needle (22G) with image guidance. 3 Fine Needle Aspiration (FNA) is extensively used for diagnosis of pulmonary lesions. A quick diagnosis is crucial to ensure satisfactory management. 4,5 The PSC system categories are used in terms of Bethesda system for reporting cervical and thyroid cytology.
The Papanicolaou Society of Cytopathology (PSC) developed a set of guidelines for respiratory cytology in 2016. The guidelines proposed six categories include non-diagnostic (ND), Negative for malignancy (NM), Atypical (A), Neoplastic (benign neoplasm or low-grade carcinoma) (N-B-LG), Suspicious for malignancy (SM), and malignant (M). Each diagnostic category is associated with certain amount of risk of malignancy (ROM). This categories and ROM determine the nature of the disease and management protocol. 6,7 In our present study, we applied the PSC nomenclature guidelines to analyze respiratory cytology specimens and evaluate risk of malignancy.

Materials And Methods
The study was exempt from approval by the institutional ethics committee, the study involved The FNA slides stained by Papanicolaou stain, Hematoxylin and Eosin stains. Slides were independently reviewed by two cytopathologists each having more than 10 years' experience in cytology and unaware about the original diagnosis. Each of 384 samples were assigned to one of the six categories described in standardized terminology for respiratory cytology proposed by PSC. Consensus diagnosis was made wherever there was discrepancy in diagnosis. At the time of making consensus diagnosis, the "triple diagnosis" technique was applied with correlation of clinical follow-up, imaging and cytopathologic ndings. Histologic samples were considered the gold standard and were available for 247 cases. Risk of Malignancy for each category was calculated for cases where biopsy was available for correlation.

Results
Total 384 consecutive cases had been taken. Patient's demographic data obtained. (TABLE 1) (TABLE 3) The ROM for each diagnostic category were 24.24% for ND, 20% for negative, 100% atypical, 0% for benign, 64.7% for suspicious, 85.81% for malignant. (TABLE 5) Overall, there was lesser incidence of atypical category and more speci c diagnoses were possible on cytology in non-malignant category.
Sensitivity and speci city for malignant cases were 99.33% and 100% respectively.

Discussion
Over the last decade, many guidelines have been laid for reporting cytology specimens by providing de nitions and cytology criteria. These categories help in further management decisions, 3 However, until recently there have been no extensively used reporting guidelines recommended for pulmonary cytology, that may provide a path towards patient management. A new guideline, PSC system of reporting lung cytology was issued that predicted ROM in pulmonary lesions. 4,8 A chief objective was to accomplish a uniform diagnostic terminology among cytopathologists. 9,10,11 In our study, 384 cases obtained by image guided FNA of respiratory lesions were examined. The ROM for Non-diagnostic (ND) category in PSC is 40%, but in our study ROM for ND category is comparatively less which is 26.5%. The lack of diagnostic features on cytology could probably be because of lesional heterogeneity in cellularity, difference in the expertise and sampling limitation. Further, our institution being a tertiary care center, chances of missing a malignancy will be of signi cant chance if a repeat cytology or biopsy is not performed when there is a presence of any clinically or radiologically suspicious lesion.
The category Negative (for malignancy) (NM) has also acknowledged a signi cant risk of malignancy (10%), which could be due to heterogenous nature of the lesion and limitation of sampling of cases.
These results, however, is much less than that mentioned in the PSC guidelines (24%-43%). 12 This category included four cases initially. First case was called as granulomatous in ammation on cytology owing to presence of multinucleated giant cells and in ammatory cells. Biopsy showed adenocarcinoma as well as multinucleated histiocytes. Other case was initially called as abscess on cytology which turned out as adenocarcinoma with areas of necrosis. The cytology showed abundance of neutrophils that obscured the tumor cell sheets and clusters. This exempli ed the pitfall in cytology and close observation to rule out malignancy. Third case was called as pneumonia on cytology while the biopsy too showed necrosis with a small cluster of tumor cells. The fourth case was synovial carcinoma on histology mimicked as epithelioid histocytes on cytology. All the other cases were infections such as granulomatous infection, fungal or pneumonia. (Fig 1, 2A,2B) FNA is very useful tool to early diagnosis of granulomatous and fungal infections. 13 In our study, majority of the cases in this category was negative for malignancy with speci c diagnosis being rendered as suppurative, granulomatous or fungal infections. This proves that cytology is a valuable tool to pick many infections that mimic malignancy as agreed by other authors. 10 Category-III (Atypical) has 100% ROM in our study. We found 3 cases in this category, all of which turned out to be adenocarcinoma on histopathological examination. (Fig 2C, 2D) Category-IV (Neoplasm-benign, undetermined malignant potential) was not comprised of any cases in our study. This category includes benign and tumors with borderline nature.
Category V-Suspicious for malignancy has shown ROM (64.7%), which is concordance to implied ROM for suspicious cases as per PSC guidelines. 12 The ROM for malignant category (Category VI) was 85.81%, which is parallel to the percentage ROM mentioned in PSC guidelines viz 77-100%. 12,14 This category includes adenocarcinoma, squamous cell carcinoma, NSCLC-NOS, pleomorphic carcinoma, neuroendocrine tumors. One of the cases of malignant category in our study is pulmonary blastoma which is very rare primary lung malignancy with worst prognosis with early age presentation. 15 Cytological smears were highly cellular and showed biphasic pattern, and based on clinical and radiological correlation the diagnosis was rendered. 16 When categories V and VI were combined, sensitivity and speci city for malignancy by cytology was 99.33% and 100% respectively. Many similar studies had sensitivity and speci city in comparable range viz., 96 -100% and 69 -97% respectively. 17,18,19 Similar to other reporting systems in cytology, such as the Bethesda system for thyroid cytology 20 and PSC for pancreatobiliary system, 13,21,22 the PSC system for reporting lung cytology was designated for improved communication between clinicians and pathologists that play a practical role in patient's treatment. According to the guidelines of PSC, in the presence of pulmonary nodule, mass or cyst on radiological study and negative ndings on cytology, the category designated is non-diagnostic category.
This implies that, in such a scenario one cannot ignore malignancy may require further evaluation that suggests a diagnosis. 12 The inter observer agreement with regard to reproducibility of various categories was not evaluated in this study which points towards a limitation of the study.

Conclusion
The study throws a light on the sensitivity and speci city of the ne needle aspiration cytology of lung lesions and is in line with the PSC guidelines. Our study substantiates that use of PSC guidelines improves the overall reporting of respiratory cytology due to the use of standardized terminology. Further studies on cytology of lung lesions obtained by other methods may further substantiate the reproducibility of this method of reporting. The raw data of this paper are available upon reasonable request to the corresponding author.

Competing interests:
The authors declare that they have no competing interests.