The value of rapid on-site evaluation during conventional and endobronchial ultrasound needle aspiration in the diagnosis of mediastinal lymphadenopathy and lung cancer

Background The endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has revolutionized pulmonology by identifying cancer spread in lung cancer patients. It is now used for diagnosing sarcoidosis, tuberculosis, and lymphoma. Rapid onsite evaluation (ROSE) is a crucial tool for pathologists, assisting in sample adequacy, accuracy, and prompt decision-making. This study aimed to evaluate the efficacy of ROSE in identifying mediastinal lymphadenopathy and lung cancer during EBUS-TBNA and cTBNA. Methods Our research was a prospective study in which we examined sixty cases that were separated into two groups of similar size. The rapid onsite evaluation group and the non-rapid onsite evaluation group were subjected to either Conventional TBNA or endobronchial ultrasound-guided transbronchial needle aspiration. Results The total diagnostic yield of conventional and endobronchial ultrasound-guided transbronchial needle aspiration in both groups was 83.3% (50/60 cases). Twenty-eight cases (46.7%) were positive for malignancy, 22 cases (36.6%) were positive for benign lesions, and 10 cases (16.6%) were not conclusive. Regarding diagnostic accuracy, it was greater in the rapid onsite evaluation group than in the non-rapid onsite evaluation group (100% in the rapid onsite evaluation group vs. 66.7% in the non-ROSE group). Conclusions Rapid onsite evaluation during conventional or endobronchial ultrasound-guided transbronchial needle aspiration improves diagnostic accuracy of mediastinal lesions by excluding suspicious or nondiagnostic specimens and can reduce unnecessary punctures or eliminate the need for additional bronchoscopy procedures when reaching preliminary diagnosis.


Background
Endobronchial ultrasound-guided transbronchial needle aspiration is a minimally invasive procedure that enables real-time TBNA with the use of ultrasound imaging.The primary purpose of its development was for the staging of lymph nodes in patients with non-small cell lung cancer.However, its application has now expanded to include the diagnosis of mediastinal lymphadenopathy and lung masses [1].
ROSE throughout EBUS-TBNA/cTBNA permits evaluation of the adequacy of the samples and material sufficiency for definitive histopathological diagnosis [2].
The purpose of the research was to assess the value of rapid onsite evaluation throughout EBUS-TBNA and cTBNA in the diagnosis of mediastinal lymphadenopathy and lung cancer.

Methods
This prospective research was performed on 60 cases with undiagnosed mediastinal lymphadenopathy or centrally located lung tumors who were referred to the bronchoscopy unit at Ain Shams University Hospital during the period between September 2021 and September 2023.

Inclusion criteria
1.An informed written consent was obtained from the cases before enrollment in the research.2. Patients referred for EBUS or cTBNA with either undiagnosed mediastinal lymphadenopathy or peribronchial mass.
All patients were subjected to the following: 1. Routine laboratory tests:CBC, bleeding profile ,kidney function and liver function Tests.2. Radiological assessment: CT chest with contrast or PET Scan. 3. Algorithmic approach: All patients underwent Transbronchial needle aspiration either with EBUS bronchoscopy (EB-19-J10U; ARIETTA V60, ALOKA, HITACHI Ltd., Tokyo, Japan) or Conventional method using flexible bronchoscopy according to following algorithm (Fig. 1) [3]: Then, the studied patients were divided into two groups: each one 30 patients.

The procedure
The patient was assessed, and vital data were recorded (noninvasive blood pressure, heart rate, and pulse oximetry).
Fig. 1 Algorithmic sampling approach for mediastinal lymph node enlargement either with cTBNA or EBUS-TBNA I. Anesthesia: the technique was done under general anesthesia.II.Bronchoscopy technique: prior to performing EBUS, the flexible bronchoscope is introduced either through a laryngeal mask or an endotracheal tube.
The initial airway examination enables a full and direct assessment of all bronchial segments and subsegments, and the removal of secretions through aspiration.Subsequently, the flexible bronchoscope was extracted, and the endobronchial ultrasoundguided bronchoscope was inserted into the airway.
Regarding the size and location of LN as mentioned before, The sampling technique would be either by EBUS-TBNA or by Conventional methods using flexible bronchoscopy (EBUS-assisted TBNA).
In EBUS-TBNA (ECHO-HD-22-EBUS-P), once the targeted lymph node was detected and measured, the needle advanced through the channel of the bronchoscope.Subsequently, following puncturing the lymph node, the stylet was removed and then an auto-aspiration syringe was applied.To and fro movements were performed 15-20 times.Average 3 to 5 punctures were described as a standard number in the research protocol from each lymph node, and the exact location within the lymph node that should be biopsied.
In conventional TBNA after localization of the exact site of LN by EBUS scope, we removed the EBUS scope and then reinserted the conventional bronchoscope (EBUS-assisted TBNA).
In contrast to EBUS, which allows for the attachment of a needle to the bronchoscope, conventional TBNA does not provide a designated location for needle fixation.We ensured the precise insertion of the needle into the intended spot, followed by securing the needle end and attaching suction.We had punctures ranging from three to six times.
Specimen handling: for patients assignedto ROSE: following each pass, the needle was withdrawn, and a small amount of material was applied to a slide for preparation using rapid hematoxylin and eosin stain that used for the ROSE technique.
Then, ROSE preliminary diagnoses were categorized into the following [4]: a Inadequate: i. Material present on slides is not sufficient and consistent with a definitive final diagnosis.
b Adequate: i. Material present on the slides is diagnostic of some process.
c Atypical: i. Cytologic features are present that may or not be associated with a pathologic process.
d Suspicious: i. Cytologic features suspicious but not definitive for malignancy, are present.
e Malignant: i. Cytologic features present on the slides are consistent with a malignant [5].
In cases assigned to the non-rapid onsite evaluation group, all cytologic specimens that were smeared were preserved in 95% alcohol for cytologic analysis.
III.All samples were sent for cytological and histopathological examination.

Statistical analysis
The

Results
The age of cases varied from 15 to 81 years with mean age ± SD being 58.37 ± 7.63 years.More than half of cases (56.7%) were males while 26 (43.3%) were women with men to women ratio was 1.31:1.Hypertension was the most frequent chronic disease found in 25% of cases followed by DM in 13.3% of cases then COPD in 3.3% of cases.
Regarding site and type of lesions in our study, station 7 was the most frequent site found in 43.3% of cases followed by 4R in 23.3% of cases then peri-bronchial mass in 15% of cases as described in Table 1.
Regarding ROSE results.Non-caseating granuloma (Fig. 2) was the most frequent benign lesion found in 30% of cases while adenocarcinoma was the most frequent malignant lesion found in 30% of cases while 3 cases showed suspicious lesions with no atypical cases as shown in Table 2 As regards malignant lesions, 14 (23.3%) had adenocarcinoma (Fig. 3), 4 (6.7%) had lymphoma (Fig. 4), three (6.3%)large cell carcinoma, and three (6.3%)small cell carcinoma out of the 15 cases with a positive malignancy diagnosis.total diagnostic yield in the rapid onsite evaluation group was 100% while 66.7% in the non-rapid onsite evaluation group 0.63.3% and 30% of cases were ultimately diagnosed with a malignancy in the rapid onsite evaluation and non-rapid onsite evaluation groups respectively and benign etiologies accounted for 36.7% of cases in both groups.Significant variations were detected among both groups regarding diagnostic yield as seen in Table 3 Based on the final diagnosis as a reference standard, the initial rose diagnosis identified malignant lesions in 19 patients (true positives) and benign lesions in 11 patients (true negatives).We found that the initial rose diagnosis had overall specificity, sensitivity, and diagnostic accuracy of 100%, 100%, and 100% respectively.The negative predictive value was 100%, while the positive predictive value was 100%, as illustrated in Table 4.

Fig. 4 Features of lymphoma on rapid H&E stain
Also in Table 6, regarding the overall diagnostic yield of cTBNA versus EBUS-TBNA there is increased sensitivity of (cTBNA) in our study to reach 61.5% while EBUS TBNA was 70.6% in the non-ROSE group.

Discussion
The present research was performed on 60 cases (34 males and 26 females with a mean age being 58.37 years.This was consistent with the research performed by Harris et al. [6] who enrolled twenty-six cases (14 men, 12 women) with a mean age of 61.4 ± 22 years, and similar to the Puriet al. [7] study registered 52 cases (thirty 33 men, nineteen women), with a mean age of 64 ± 12 years.Furthermore, the most common sampled lesions was the subcarinal lymph node (station 7) followed by 4R in 23.3% of cases then peri-bronchial mass in 15% of cases This is similar to Kassirian et al. [8].research and the Dhooria et al. [9] research.However, it was unlike the Mohan et al. [10], Casalet al. [11], and YarmusL.B et al. [12], studies in which the most sampled lymph node was the right paratracheal (4R) lymph node.
In the current study, the total diagnostic yield of Conventional and Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration in both groups was 83.3% (50/60 cases).Twenty-eight cases (46.7%) were positive for malignancy, 22 cases (36.6%) were positive for benign lesions, and 10 cases (16.6%) were not conclusive.The diagnostic yield and sample adequacy rates closely match the figures provided in the AQuIRE registry, with a rate of 50% in sarcoidosis cases and 90% in malignancy cases.This registry has given data from real-world scenarios Ost et al. [13].This was in alignment with the study performed by Dhooriaet al. [9] which presented that the diagnostic yield of endobronchial ultrasound-guided transbronchial needle aspiration was 61.2%.It was also like the research performed by Tremblay et al. [14] which proved that the diagnostic yield was 83.3%.This study was also in concordance with the research done by Adams et al. [15] which presented that the diagnostic yield was 88%, and another study by Herthet al. [16] which proved that the diagnostic yield was 71%.
Regarding diagnostic accuracy, the rapid onsite evaluation group had a greater accuracy rate compared to the non-rapid onsite evaluation group (100% in the rapid onsite evaluation group versus 66.7% in the non-rapid onsite evaluation group, p = 0.10).Similar findings were stated by Cardoso et al. [17], who demonstrated a greater diagnostic success rate in the rapid onsite evaluation group (93%) compared to the non-rapid onsite evaluation group (77%).
Guo et al. [18] conducted a meta-analysis and observed a tendency for improved outcomes with rapid onsite evaluation.However, the presence of heterogeneity among the studies hindered the achievement of satisfactory outcomes.However, Okiet al. [19] conducted a randomized trial while Griffinet al. [20] conducted a retrospective investigation, both of which concluded that rapid onsite evaluation did not have a significant impact on the diagnostic yield of endobronchial ultrasound-guided transbronchial needle aspiration.
Regarding the number of punctures, there is no disparity between both groups.Griffin et al. found that rapid onsite evaluation throughout endobronchial ultrasound-guided transbronchial needle aspiration did not reduce the number of lesions sampled per case, which aligns with our findings.Conversely, Oki et al. andBediwy et al. [21] proposed that rapid onsite evaluation decreased the number of punctures required for each lesion or the number of lesions that needed to be aspirated.Furthermore, Cardoso et al. observed that the utilization of rapid onsite evaluation throughout EBUS-TBNA resulted in a reduction in the number of punctures required per procedure.Nevertheless, the rise in punctures within the non-rapid onsite evaluation group did not correlate with a high rate of complications.In our work, we addressed the issue surrounding literature by employing both Conventional and EBUS TBNA procedures, with various needle diameters to acquire smears.Typically, we adjust the procedure by altering the location, depth, or angle of the puncture based on the ROSE results.As a result, the rapid onsite evaluation group suffers a greater percentage of punctures.
The average time of bronchoscopy was longer in the rapid onsite evaluation group.This finding aligns with the research carried out by Oki et al., who concluded that using ROSE did not result in a reduction of bronchoscopy time due to the time-consuming process of preparing and reviewing slides for ROSE.
Out of the thirty cases in the ROSE group, only three cases were found to be suspicious, while the remaining twenty-seven cases were strongly indicative of either malignant or benign lesions, as previously described.Guo et al. confirmed the benefit of rapid onsite evaluation in their research.They found that the ROSE group had a lower percentage of suspicious results (8.7%) compared to the non-rapid onsite evaluation group (14.6%) (P = 0.038).Additionally, the use of rapid onsite evaluation led to a higher diagnostic yield of pathologic samples, with a success rate of 90.5% compared to 81.2% in the non-rapid onsite evaluation group (P = 0.003).The diagnostic precision of rapid onsite evaluation in determining the presence or absence of malignancy in the final pathological diagnosis was assessed on a per-lesion base.The initial rose diagnosis correctly diagnosed malignant lesions in nineteen cases (true positives) and benign lesions in eleven cases (true negatives), resulting in a sensitivity and specificity of 100% each.The findings were consistent with the research conducted by Bediwyet al. which documented a specificity of 94.12%, a sensitivity of 94.87%, and a diagnostic accuracy of 94.57%.Yasufukuet al. [22] found that the sensitivity and specificity were 95.7% and 100%, correspondingly, which indicates a better level of accuracy.Ye et al. [23] reported a sensitivity of 95% and a specificity of 100%.Al Sharifet al. [24] and Parmaksiz et al. [25] reported an 89.2% sensitivity and a 100% specificity, correspondingly.Guo et al. found a high level of agreement (98.6%) among the onsite observations and the final pathology diagnosis.

Table 1
Distribution of studied patients regarding type and site of lesion Fig. 2 Non-caseating granuloma as seen through a microscope on a glass slide during rapid H&E stain

Table 2
Preliminary ROSE results among studied patients

Table 3
Diagnostic yield among the studied population

Table 4
Accuracy measures of initial rose diagnosis in relation to final diagnosis

Table 5
Comparison between the two studied groups regarding number of punctures and bronchoscopy duration

Table 6
Comparison of the diagnostic yield of cTBNA versus EBUS-TBNA in the non-ROSE group