記住我
Abstract
The incidence of lung cancer has been increasing in the recent years. Bronchial cytology using Papanicolaou society of cytopathology (PSC) system is an effective method for triaging patients. The present study attempts to evaluate the bronchial cytological diagnosis with histopathological correlation of lung lesions. Aims: i. To study the cytological features of lung lesions. ii. To assess the sensitivity, specificity, and diagnostic accuracy of bronchial cytology of lung lesions in comparison with histopathology. Settings and Design: Prospective study at the tertiary care hospital. Methods and Material: It included 63 cases of lung lesions, evaluated using the PSC system for reporting respiratory cytology. The cytological diagnosis was correlated with the final histopathological diagnosis. The study was conducted between January 2019 and June 2020. Statistical Analysis Used: SPSS 20.0 software. Results: The sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of bronchial cytology was 60%, 89%, 90%, 58.62%, and 71.42%, respectively. Conclusions: Bronchial cytology including bronchial wash, bronchial brush, endobronchial ultrasound/transbronchial needle aspiration, and computerized tomography-guided fine needle aspiration cytology can be used to increase the sensitivity and specificity for definitive diagnosis and better management.
Keywords: BAL, brush, EBUS/TBNA, histopathology
How to cite this article: Introduction 
Lung cancer is currently the most frequent and common cause of cancer-related mortality worldwide.[1] It is the most frequent malignancy in industrialized nations.[2] The increasing incidence could be due to smoking, change in lifestyle, increase environmental pollution, and also the availability of different modern diagnostic modalities to detect lung cancer.[1]
Lung cancer is the most significant worldwide health problem and contribute to 30% of male and 26% percent of female cancer-related deaths and approximately 1.38 million people die of lung cancer each year.[3] Respiratory cytology has assumed a primary diagnostic place in evaluation of any patient with suspected lung lesion in which morphological confirmation is indicated.[4] To treat the disease successfully, it should be diagnosed at earliest stage. For early diagnosis, different diagnostic modalities are available that includes radiology, bronchoscopy and bronchial biopsy, and brushing and washing cytology. As each technique has advantages and disadvantages, it is not possible to perform all techniques in each patient. However, their combined use yields the best results.[1],[5]
Bronchoscopy-aided biopsy cannot visualize early malignancy nor can be performed in more peripheral sites. It cannot be done in patients with risk of hemorrhage and in tumors that have lepidic growth. So alternative methods for diagnosis are required. Hence, bronchial cytology will compliment tissue biopsy in lung lesions.[6]
The bronchial brushing is a safer technique and malignancy can be readily recognized and typed. World literature depicts the sensitivity, specificity, and diagnostic accuracy of bronchial cytology as 78.70%, 87.42%, and 79.35%, respectively, in lung lesions.[7],[8] Therefore, we are conducting this study to analyze the cytological features of lung lesions and to determine the accuracy of bronchial cytological diagnosis of lung lesions in comparison with histopathology.
Subjects and Methods This was a prospective time bound study conducted in department of pathology in a Hospital from January 2019 to June 2020. Ethical clearance was taken from institutional review committee. This study included 63 cases of lung lesions. Detailed clinical examination, chest X-ray, and routine laboratory investigations were done in all the cases of clinically suspected lung lesions. Prior informed consent was taken from the patients for bronchial cytological examination. The samples were obtained by fiber-optic bronchoscope by the chest physician at department of Chest and Tuberculosis, at the hospital. Bronchial brushings specimen of two to three brushings by disposable bronchial brush, was collected and smeared on the glass slides. The smears of both brush and wash cytology were fixed with methanol and few slides were air dried. Methanol-fixed slides were stained with Papanicolaou and air-dried slides with Leishman stain. Cytological smears were classified into six categories according to Papanicolaou society of cytopathology (PSC) system for reporting respiratory cytology. Bronchial biopsy specimens were collected by the bronchoscope. The specimens were fixed with 10% formalin for one day and sections were prepared and stained with hematoxylin & eosin stain. On histopathology, lesions were classified into benign and malignant based upon the histological findings. Histopathological diagnosis was considered as the gold standard. Histopathological features are studied and compared with cytological interpretation. True positive (TP), true negative (TN), false positive (FP), and false negative (FN) are all calculated. Then sensitivity, specificity, diagnostic accuracy, positive predictive value (PPV), and negative predictive value (NPV) are interpreted with the help of following calculations.
Calculations:
Sensitivity = TP/TP + FN × 100.
Specificity = TN/TN + FP × 100.
PPV = TP/TP + FP × 100.
NPV = TN/TN + FN × 100.
Diagnostic accuracy = TP + TN/Total number of cases.
TP = Positive result for neoplasm on cytology with histopathological confirmation.
FP = Positive result for neoplasm on cytology with a benign lesion on histopathology.
TN = Negative result for neoplasm with subsequent final histopathological confirmation.
FN = Negative result for neoplasm on cytology with a neoplastic lesion on histopathology.
Results Among the 63 cases of solitary lung lesions, males were predominantly affected with a Male: Female ratio of 2.7:1. The mean age was 57 years. The cytological diagnosis was categorized into six categories of PSC system and cytological diagnosis was given as shown in [Table 1], and [Figure 1] and [Figure 2].
Figure 1: (a) Cytological smear of category II (Negative for malignancy) - Granuloma in BAL displaying clusters of epithelioid histiocytes, scattered lymphocytes and fibrous tissue fragments. (Leishman stain, ×400), (b) Cytological smear of category II (Negative for malignancy) - Langhans giant cell in BAL in a case of tuberculosis. (Papanicolaou stain, ×400). (c) Cytological smear of category II (Negative for malignancy) - Mucormycosis in BAL displaying thick walled, wide based, aseptate fungal hyphae with right angle branching. (Papanicolaou stain, ×400), (d) Cytological smear of category III (Atypical) - Clusters of atypical cells in BAL displaying anisonucleosis, hyperchromatic nucleus, coarse chromatin, prominent nucleoli, overlapping, overcrowding and a moderate amount of dyskeratotic cytoplasm. (Leishman stain, ×400)
Figure 2: (a and b) Cytological smear of category VI (Malignancy) - Adenocarcinoma in BAL displaying tumor cells arranged in tridimensional clusters and acini having high N: C ratio, hyperchomatic nucleus, coarse chromatin, prominent nucleoli, cytoplasmic vacuoles and signet ring change. (Leishman stain, x400 and Papanicolaou stain, ×400), (c and d) Cytological smear of Category VI (Malignancy) – Squamous cell carcinoma in BAL displaying pleomorphic squamous cells with high N: C ratio, anisonucleosis, hyperchomatic nucleus, irregular nuclear margins, focal dyskeratosis and tadpoling of cytoplasm. (Leishman stain, ×400 and Papanicolaou stain, ×400)On histopathological examination, 33.33% (21 cases) were found to be nonneoplastic, and 66.67% (42 cases) were neoplastic lesions and histopathological diagnosis as shown in [Table 2] and [Figure 3].
Figure 3: Histopathological examination (a) Adenocarcinoma displaying tumor cells arranged in atypical glands having high N: C ratio, hyperchomatic nucleus, coarse chromatin and cytoplasmic vacuoles. (Hematoxylin and Eosin × 200), (b) Poorly differentiated squamous cell carcinoma displaying pseudostratified ciliated epithelium and pleomorphic tumor cells infiltrating the subepithelium in solid sheets having high N: C ratio, anisonucleosis, hyperchomatic nucleus, irregular nuclear margins, dyskeratosis and pulled out cytoplasm. (Hematoxylin and Eosin × 200), (c) Small cell carcinoma displaying pleomorphic small round cells arranged in diffuse sheets, having high N: C ratio, hyperchromatic nucleus and scant cytoplasm. (Hematoxylin and Eosin × 200), (d) Malignant spindle cell tumor displaying pleomorphic spindle cells arranged in intersecting fascicles having hyperchromatic nucleus, anisonucleosis and dense eosinophilic cytoplasm. (Hematoxylin and Eosin × 200)Correlation of bronchial brushing cytology and biopsy
Bronchial cytological diagnosis according to PSC system category was correlated with the histopathological diagnosis as shown in [Table 3]. Nonneoplastic lesions were noted in 46.03% (29 cases) on cytology. Out of these 29 cases, 17 cases remained nonneoplastic, and 12 cases turned out to be malignant on histopathological evaluation. Out of these 12 cases, nine were diagnosed as non small-cell carcinoma, two were diagnosed as small-cell carcinoma, and one case was diagnosed as mixed small- and large-cell carcinoma. These 12 cases were considered to be FN [Figure 4]. In our study, the cytohistological concordance for the nonneoplastic group was 58.6%. On cytological examination, 20 cases (31.74%) were neoplastic, out of which 18 cases remained neoplastic and two cases turned out to be nonneoplastic on histopathological evaluation. Out of these two cases, one case was diagnosed as inflammatory lesion and other was diagnosed as granulomatous lesion. These two cases were considered as FP [Figure 5]. Cytological concordance for neoplastic group was 90%.
Table 3: Correlation of the PSC categories and histopathological diagnosis
Figure 4: Cytological smear of false negative cases - (a) Displaying sheets of only neutrophils. These features can be seen if a malignant tumor in associated with secondary infection or pneumonia. (Papanicolaou stain, ×400). (b) Displaying extensive hemorrhage and few benign endobronchial cells. These features can be seen if a patient has hemoptysis associated with malignancy and due to sampling error where no malignant cells are seen in the smear (Papanicolaou stain, ×400)
Figure 5: Cytological smear of false positive case - Displaying extensive squamous metaplasia with mild nuclear atypical having coarse chromatin and high NC ratio mistaken as squamous cell carcinoma (Leishman stain, ×400)The diagnostic accuracy of bronchial cytological examination was seen in 71.42% of cases. The discrepancy rate between PSC system for reporting respiratory cytology and histological diagnosis was 28.58%. [Table 4] shows the sensitivity, specificity, PPV, NPV, and diagnostic accuracy of bronchial cytology in present study. The histopathological results were proved to have significant difference with respect to the bronchial washing and the bronchial brushing cytology with P value <0.001, which implies the significant difference in the two methods.
Discussion There has been an increasing incidence of lung lesions in recent years. Patients with lung lesions account for a major work load among the clinicians as well as pathologists. Both patients and physicians prefer using simple investigations that are cost effective and have shorter hospital stay.[9] Bronchial cytology (Brush and wash) can be done as an initial screening investigation. The pulmonary cytopathology practice requires communication and collaboration among pathologists and clinicians for further management. Thus, the standardized guidelines and nomenclature have been introduced using the PSC guidelines for reporting respiratory cytology.[10],[11]
This study aims to study the cytological features of lung lesions and to assess the sensitivity, specificity, and diagnostic accuracy of bronchial cytology of lung lesions in comparison with histopathology. In our study, the age of the patients ranged from 25 to 85 years. The mean age of presentation was 57.60 years, with most of the patients presenting between 56 to 65 years. The age distribution was comparable to the studies done by Behura et al.,[4] B Sultana et al.[12], Giti et al.,[13] and Sareen et al.[14] The male-to-female ratio in this study was 2.7:1 that was comparable with the study done by Choudhury et al.[15]
In our study, the cause of FN resulting in 12 cases was due to sampling error in smaller neoplastic lesion, peripheral tumor location, inadequate exfoliation of malignant cells, and abundant inflammation masking the tumor cells. Similar observations regarding FN were done by Feeley et al.[16] and Nodit et al.[17] Feeley et al.[16] reported 57 FN results, majority of which were due to peripheral tumor location. Nodit et al.[17] found 32 cases as FN, in which the sampling error was the major cause.[17] FN results may also be due to confounding inflammation, inadequate exfoliation of malignant cells, interpretive errors and poor distribution of bronchoalveolar lavage (BAL) specimens. Since BAL cytology relies mainly on exfoliated cells from malignant tumor, but for unknown reasons some lung cancers do not exfoliate diagnostic cells regardless of the number of specimens collected.[18] Hence, better communication with clinicians for performing computerized tomography (CT)-guided fine needle aspiration cytology (FNAC)/biopsy when peripheral lesions and repeat bronchial cytology/endobronchial ultrasound (EBUS) – transbronchial needle aspiration (TBNA)/biopsy in cases of suspected endobronchial/parenchymal mass lesion.[13]
In our study, the cause of FP (two cases) was due to squamous metaplasia with atypia. This was in accordance with the study conducted by Raiza et al.[1] who reported one FP case in their study. False positivity may occur due to misinterpretation of cytological findings by cytopathologists due to reactive atypical changes in epithelioid cells and histiocytes in inflammatory conditions (example tuberculosis, pneumonia etc.) or squamous metaplasia with atypia (commonly seen in chronic smokers) as seen in our case. These FP results may have serious consequences in patients, for whom biopsy is not possible due to anatomical obstruction or risk of hemorrhage. Hence it is advisable to under diagnose atypical/suspicious smears.[1]
Cytohistological correlation was seen in 71.42% of cases. The discrepancy rate between PSC categories and histological diagnosis was 28.58%. In the present study, the sensitivity was 60% and the specificity was 89%. The sensitivity in our study is comparable with the study done by Razia et al.[1] (72.72%) where bronchial brushing was combined with bronchial wash. The sensitivity in our study is higher than the study done by Tayal et al.[19] (47.83%) where only bronchial wash was used and lower than the study done by Gaur et al.[20] (91.7%) where brush cytology was combined with TBNA. The specificity in our study was 89%, which was comparable with the study done by Razia et al.[1] (92.85%) and was higher than the study done by Gaur et al.[20] (47.83%) where brush cytology was combined with TBNA. PPV in our study was 90% which was comparable with the study done by Tayal and Bhale.[19] and Vyas et al.[2] The NPV in our study was 58.62% which was comparable with Choudhury et al.[5] The diagnostic accuracy in our study was 71.42%, which was comparable with the study done by Razia et al.[1] (80.5%) and Tayal et al.[19]
The limitation in the present study was low sensitivity (60%) which was due to a relatively higher number of FN. The FN rate in our study was 19%. The most frequent cause of FN cytological diagnosis was sampling error, parenchymal lesions, peripheral tumor location, and abundant inflammation. In our study, the probable explanation of low sensitivity is the use of only BAL cytology for majority of the cases (85%), and BAL with brush cytology only in few cases (15%). BAL yields less tumor cells due to peripheral tumor location, abundant inflammation, and presence of degenerative cells. Hence, to increase the sensitivity of bronchial cytological examination, BAL should be combined with bronchial brush and TBNA (EBUS guided).[18],[21]
Conclusion Bronchial cytology includes bronchial wash (BAL) and brush cytology. Since, it has low sensitivity, further evaluation should be done using repeat bronchial cytology, EBUS/TBNA, or CT-guided FNAC/biopsy especially if there is clinical/radiological suspicion of malignancy for definitive diagnosis and management.
The PSC system for reporting respiratory cytology is an extremely useful standardized system that categorizes the lung lesions among six-tier system. It improves communication in a multidisciplinary setup of pathologists and clinicians, thereby minimizing the inter laboratory variation, leading to more consistent management approaches. It has proven to be an effective guide for clinical management of lung lesions. The BAL and bronchial brush cytology has high specificity and diagnostic accuracy but low sensitivity, thereby concluding that bronchial cytology (BAL + Brush) is not a good screening tool for malignancy. Hence, this should be accompanied with EBUS/TBNA for screening and diagnosing lung cancer as it gives better cellularity, nuclear characteristics, cellular preservation, chromatin details, and nucleoli. Always respiratory cytological examination (BAL, brush, EBUS/TBNA, and CT-guided FNAC) should be accompanied by cell block/biopsy for confirmation and ancillary investigation like immunohistochemistry and molecular studies.
Key messages
Bronchial cytology (BAL + Brush) is not a good screening tool for malignancy; hence, this should be accompanied with endobronchial ultrasound/transbronchial needle aspiration for screening and diagnosing lung cancer as it gives better cellularity, nuclear characteristics, cellular preservation, chromatin details, and nucleoli.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Acknowledgements
At the outset, I would like to express my sincere gratitude to Dr. Sunil Kumar Y. Additional Professor, Department of Pathology, who has been the guiding light since the inception of this research work.
I would like to wholeheartedly thank my postgraduate colleagues Dr. Ritu Asnani and Dr. Divya Joshi for their contribution and persistent support.
I am thankful to the patients, without whom this study would never be possible.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References 
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/joc.joc_44_22
留言 (0)