Traveling with ROSE in EBUS-TBNA - Experience from a Tertiary Care Hospital in South India.

AIMS AND
OBJECTIVES: Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has emerged as a precise modality for tissue sampling of mediastinal and hilar lesions adjacent to the proximal airway. This study aims to determine the diagnostic efficacy, sensitivity, and specificity of rapid-on site evaluation (ROSE) in EBUS-TBNA.
MATERIALS AND METHODS: This is a retrospective study that included 100 patients who underwent EBUS-TBNA of paratracheal and mediastinal lymph nodes in a tertiary care hospital in South India between March 2018 and March 2020. After the procedure, the diagnostic yield from the nodes sampled was transferred to slides that were stained with rapid hematoxylin and eosin (H and E), and then onsite evaluation was done. The tissue derived was also processed for histopathologic examination in all cases. ROSE was performed to assess sample adequacy and to arrive at a preliminary diagnosis. In patients suspected of tuberculosis, the sample was collected for GeneXpert evaluation as well.
RESULTS: Of the 100 cases studied, 51 were males and 49 were females. The age distribution was between 3 and 78 years. Forty-seven cases were diagnosed as granulomatous lymphadenitis, 13 as metastatic malignancies, 33 as reactive lymphadenitis, 3 as atypical cells, and 1 case was diagnosed as a cystic lesion. The diagnostic yield was not adequate for evaluation in three cases. Diagnostic yield was obtained in the first two passes where the lymph nodes were more than 2 cm in size. More diagnostic passes were required in lymph nodes less than 2 cm and those located between and adjacent to major vessels. The onsite diagnosis was correlated with the final histopathologic diagnosis.
CONCLUSION: ROSE serves as a useful adjunct to reduce procedure time and enhance sample collection and triaging, and reduces the need for further invasive testing. Copyright:

INTRODUCTION

India is a country with a high incidence and prevalence of tuberculosis (TB) as well as lung malignancy. Most of these patients present with mediastinal and hilar lymphadenopathy. Many novel techniques are available to sample these lymph nodes, including blind conventional bronchoscopic transbronchial needle aspiration (TBNA), mediastinoscopy, transesophageal ultrasound-guided needle aspiration, and computed tomography (CT)-guided transthoracic needle aspiration.[12] But some of these techniques have certain disadvantages, like the need for general anesthesia, low yield, and poor accessibility. The diagnostic yield is approximately 42%-55% in conventional TBNA and mediastinoscopy.[34]

Endobronchial ultrasound-guided TBNA (EBUS-TBNA) is a technique that allows performing on “'real-time”' TBNA, under ultrasound-guided imaging. EBUS-TBNA is performed as an outpatient procedure under local anesthesia and conscious sedation. The Doppler mode in EBUS-TBNA allows vascular structures to be visualized better and also helps in the accurate localization of lymph node stations. After the procedure, the clinicians are not sure of the adequacy of the material sampled, including the vital question of “Have we sampled the lesion?” There comes the role of cytopathologists at the procedural site. They can take a smear of the sampled tissue and do a rapid stain and look at it under the microscope. They can confirm that adequate material is sampled from the lesion and most of the time give a preliminary diagnosis. This is called rapid onsite evaluation (ROSE). ROSE plays an important role in confirming sample adequacy and this prevents the need for the re-procedure. The added benefit is the preliminary diagnosis, but this has to be confirmed by the final diagnosis after processing the whole tissue. Another vital role of ROSE is to appropriately triage the sample depending upon the finding; to name a few, sending the material for GeneXpert for TB testing in cases of granulomas and molecular testing in cases of non-small cell malignancies.

The variability in diagnostic yield depends on the etiologies of enlarged lymph nodes, the type of needle used, the number and size of the lymph nodes sampled, the number of passes obtained from each lymph node station, the availability of rapid onsite cytological evaluation (ROSE), and the expertise of the operator.[5] The study aims to determine the diagnostic efficacy, sensitivity, and specificity of ROSE in EBUS-TBNA. The present study analyzes the diagnostic yield with lymph node size and the number of passes and also the usefulness of ROSE in EBUS-TBNA, correlating onsite evaluation preliminary diagnosis with the final histopathologic diagnosis.

MATERIALS AND METHODS

This is a retrospective study that included 100 patients who underwent EBUS-TBNA in a tertiary care hospital in South India between March 2018 and March 2020. Patients of all age groups with mediastinal and hilar lymphadenopathy, who underwent EBUS-TBNA, were included in this study. All the procedures were done in the bronchoscopy room by pulmonologists. After the procedure, the needle was removed from the scope and the specimen was pushed out from the needle sheath by reinserting the stylet and carefully flushing it by air or saline onto the slide. The diagnostic yield from the nodes sampled was transferred to slides and was stained with rapid hematoxylin and eosin (H and E) staining, and then onsite evaluation was done. ROSE was performed to assess sample adequacy and to arrive at a preliminary cytological diagnosis at the site of the procedure and the results were communicated to the pulmonologists within 15 minutes. The solid tissue derived from the procedure was then placed in 10% buffered formalin for histologic evaluation. The formalin-fixed, paraffin-embedded samples were stained by routine H and E staining and subjected to histopathologic examination for all cases.

The assessment for the sample adequacy was performed based on the following criteria: presence of tissue core in EBUS needle equal to or more than 2 cm or presence of malignant cells or microscopic anthracotic pigment or lymphocyte density of more than 40 per 10 high-power fields at 40× magnification.[6]

In cases with the preliminary diagnosis of epithelioid granulomas with necrosis, the aspirates from subsequent passes were sent for detection of Mycobacterium by the GeneXpert method. Sarcoidosis was diagnosed in cases with compact epithelioid cell granuloma without necrosis correlating clinically, radiologically, and serologically. The diagnosis of malignancy was obtained based on the cytological features. A diagnosis of reactive lymphadenitis was rendered in slides showing the heterogeneous population of lymphocytes and tingible body macrophages after adequate sampling.

Descriptive and inferential statistical analysis was carried out in the present study. Data were recorded in a database designed in Microsoft Office Excel. Diagnostic yield, sensitivity, specificity, positive predictive value, and negative predictive value were computed using appropriate equations.

RESULTS

Of the 100 cases studied, 51 were males and 49 were females. The age distribution was between 3 and 78 years. The demographic profile and clinical diagnosis for these cases are presented in Tables 1 and 2. Diagnostic yield was obtained in the first two passes when the lymph node was more than 2 cm in size. All levels of lymph nodes (2R, 2L, 4R, 4L, 7R, 10L, 11R, 11L, 12R, 12L) were sampled in different settings, with the commonly sampled being the subcarinal nodes. More diagnostic passes were required in lymph nodes less than 2 cm in size and those located between and adjacent to major vessels.

Table 1

Demographic characteristics

Demographic profile
Age	3 to 78, (median-46)
Male	51%
Female	49%

Table 2

Distribution of cases according to clinical diagnosis.

Clinical Diagnosis	%
Granulomatous lesion	66%
Malignancy	32%
Others	2%

A sample is labeled as diagnostic if it provides a definitive diagnosis such as TB, sarcoidosis, malignancy, and others. For uniformity of reporting, the diagnostic category are reported as nondiagnostic, negative for malignancy (which included granulomatous inflammation with or without necrosis and other reactive conditions), atypical cells, suspicious of malignant cells, or positive for malignant cells (adenocarcinoma, squamous cell carcinoma, small cell carcinoma, and others).[7] The diagnoses obtained through EBUS-TBNA are shown in Table 3.

Table 3

Diagnosis obtained through EBUS-TBNA

Diagnosis	Number of cases
Non-diagnostic	3
Cystic lesion	1
Non-Malignant
Reactive lymphadenitis	33
Granulomatous lesion
Tuberculosis	33
Sarcoidosis	14
Atypical cells	3
Malignant
Adenocarcinoma	9
Squamous cell carcinoma	3
Others (Metastasis from leiomyosarcoma)	1

Of the 100 cases, 47 were diagnosed as granulomatous lymphadenitis [Figure 1a and b], 33 as reactive lymphadenitis, 3 as atypical cells, 13 as metastatic malignancies [Figure 1c and d], and 1 case was diagnosed as a cystic lesion. The diagnostic yield was not adequate for evaluation in three cases. Of the malignant cases, non-small cell carcinoma was commonly diagnosed. The onsite diagnosis was correlated with the final histopathologic diagnosis. Table 4 shows the bronchoscopic characteristics and correlation of onsite diagnosis with the final histopathologic diagnosis.

Figure 1

(a) 18 × 18.3 mm subcarinal station 7 lymph node with central necrosis. (b) Aspirate with epithelioid histiocytes and background showing lymphocytes. (c) 10.5 × 11.3 mm level 4R heterogeneous lymphadenopathy. (d) Aspirate showing clusters of atypical cells with background showing lymphocytes, bronchial epithelial cells, and macrophages. (e) 25.7 × 25.9 mm left interlobar 11 L heterogeneous lymph node. (f) Cellular smears composed of sheets of atypical cells exhibiting marked pleomorphism

Table 4

Bronchoscopic characteristics and correlation of onsite diagnosis with final histopathological diagnosis

Diagnosis	ROSE	Final HPE	Lymphnode size (mean)	The average number of passes per patient
Granulomatous lesion	47	46	18.8-34.1mm	3.6
Reactive lymphadenitis	33	34	19.6-32.3mm	4.6
Metastatic malignancy	13	13	22.5-44.2mm	2.8
Others (cystic lesion)	1	1	75.9mm	1

Granulomatous lymphadenitis was reported in 47 cases, of which 33 were of tuberculous origin and 14 were diagnosed as sarcoidosis. GeneXpert evaluation was done in 33 cases, in which 14 cases were positive for Mycobacterium tuberculosis. Diagnosis of sarcoidosis was based on cytology, elevated acetyl choline esterase (ACE) levels, negative Mantoux test, and clinical response to treatment. Of the metastatic malignancies, nine were adenocarcinomas with a primary lung mass in all the cases, three were squamous cell carcinomas with two cases having a primary lung mass and another case having an esophageal malignancy, and one case was a metastatic leiomyosarcoma [Figure 1e and f] possibly from the uterus as the patient had a history of uterine mass that had been operated before 2 years.

The EBUS-TBNA procedure was not associated with any significant complications like pneumothorax, major vessel injury, or severe bleeding requiring hospitalization or intensive care monitoring. Bleeding was minimal in most cases, and the procedure was well tolerated by all age groups. The diagnostic yield, sensitivity, specificity, positive predictive value, and negative predictive value of the present study are presented in Table 5. A comparison of these results with those of similar studies published is given in Table 6.

Table 5

Statistical analysis of ROSE in EBUS TBNA

Parameters analyzed	Results
Diagnostic yield	97.7%
Sensitivity	94%
Specificity	100%
Positive predictive value	100%
Negative predictive value	12%

Table 6

Comparison of the results of the present study with similar studies

Diagnostic yield (%)	Sensitivity(%)	Specificity(%)	Positive predictive value(%)	Negative predictive value(%)	References
97.2	95.08	100	100	93.02	Ye T et al., [8]
74.5	81.7	100	100	22.73	Madan K et al., [9]
93	94	100	100	11	Herth FJ et al., [16]
91.02	89.55	100	100	61.11	Arun Nair et al.,[10]
97.7	94	100	100	12	Present study

DISCUSSION

The wide range of etiologies for mediastinal and hilar lymphadenopathies poses a diagnostic challenge. EBUS-TBNA has proved as a powerful weapon, being performed as an outpatient procedure under local anesthesia and conscious sedation, in diagnosing these cases. The aim of the present study is to determine the diagnostic efficacy, sensitivity, and specificity of ROSE in EBUS-TBNA.

Of the 100 cases studied, 51 were males and 49 were females with a male to female ratio of 1.04:1. The age distribution was between 3 and 78 years. The average age of the patients in our study was 48 years, which was much less as compared to a study by Chhajed et al.,[11] where the average age was reported as 66 years. In the current study, 32% of the cases were evaluated with a clinical suspicion of malignancy and the rest were evaluated for benign disorders like TB and sarcoidosis. On average, 1.8 lymph nodes were sampled per patient. The average number of passes per patient was about 3.6 for granulomatous lesions, 4.6 for reactive lymphadenitis, and 2.8 for malignancies.

The procedure was done on a day-care basis under local anesthesia and conscious sedation, which helps in reducing the cost of treatment, hospital stay, and also improves patient satisfaction. There were no significant complications associated with the procedure. We obtained sample adequacy of 96.8% in the present study, which is in concurrence with the studies of Yasafuko et al.,[2] who also reported adequacy of 96%, and Mallya et al.,[12] with the sample adequacy reported as 96.3%. A similar study by Alsharif et al.[13] reported sample adequacy of 84.3%.

Our study showed a greater preponderance of nonmalignant lesions compared to malignant etiology in mediastinal lymphadenopathy in this series. This finding is similar to the studies from other parts of India,[1415] where granulomatous lesions constituted the most common diagnosis, but is in contrast to the data from East Asia and Europe.[216] Of 33 cases, 42% of the patients diagnosed with TB had positive microbiological evidence of TB. In a study by Dhamija et al.,[14] 55% of TB cases had positive microbiological evidence. Of the malignant cases reported, the most common was a metastasis from primary lung adenocarcinoma. Parmaksiz et al.[17] also reported adenocarcinoma as the most important malignancy in their study. Categorization of the malignant cases as adenocarcinomas, squamous cell carcinomas, and others was based on the immunohistochemical staining performed in the cores obtained during the procedure. The patients diagnosed as atypical cells in ROSE showed reactive atypia in the final histopathologic diagnosis and these patients were advised for a close follow-up.

The sensitivity and specificity of ROSE were high in our study, which concurs with other similar studies [Table 6]. The diagnostic yield, sensitivity, specificity, positive predictive value, and negative predictive value are in concordance to related studies. The presence of ROSE during EBUS-TBNA aided the pulmonologist in assessing the adequacy of samples, as well as determining the need for additional sampling for ancillary immunohistochemical and molecular testing. This is of great importance in this era of personalized stratified medicine in the management of lung cancer. The presence of fewer false positives suggested that ROSE during EBUS-TBNA was useful in detecting a malignant diagnosis. Adequate sampling also aids in performing molecular analysis on the cell block or scraping the slides after destaining.[1819] In correspondence to similar studies, we concur that ROSE during TBNA significantly improves the diagnostic yield[2021] and also reduces the number of passes necessary for achieving a high diagnostic yield.[2223]

We believe that ROSE was also beneficial in facilitating the necessary ancillary tests like microbiological culture, flow cytometry, and molecular testing. The aspirate obtained can be stored in rosewell park memorial institute (RPMI) medium at 4°C for 5 days, which can be utilized to perform flow cytometry, karyotyping, and fluorescent in situ hybridization (FISH).[2425] For mutation analysis and molecular testing, material from the cell block as well as scrape cytology from the slide can be used.[2627] The common molecular testing recommended by the American Society of Clinical Oncology is EGFR, ALK, ROS, and BRAF.[28] The minimum number of cells recommended for EGFR/KRAS mutation analysis is 300 cells[29] and for ROS-1/ALK testing is 100 cells.[30] But these numbers also depend on the method used to perform molecular analysis.[3132] Programmed death ligand-1 (PDL-1) testing requires a minimum of 100 viable tumor cells on the slide or cell block.[33]

ROSE helps determine the adequacy of the aspirate for molecular testing in patients with lung cancer.[183435] EBUS-TBNA is a useful technique that has revolutionized the field of pulmonology by proving as a useful procedure in sampling lymph nodes that remain inaccessible by conventional TBNA or mediastinoscopy based on size or location. ROSE helps to act as a guide for assessing the adequacy and accuracy of the sample and aid in rapid decision-making and triaging the sample appropriately.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.