Karim MM, Ayesha S, Shahid AH, Salman M, Kumar D, Uddin Z, Rehman AU. Diagnostic accuracy and safety of endoscopic ultrasound-guided fine needle biopsy for evaluating mediastinal pathologies. World J Gastrointest Endosc 2026; 18(2): 113699 [DOI: 10.4253/wjge.v18.i2.113699]
Corresponding Author of This Article
Adeel Ur Rehman, Assistant Professor, Section of Gastroenterology, Department of Medicine, Aga Khan University, Karachi 74800, Sindh, Pakistan. adeelur.rehman@aku.edu
Research Domain of This Article
Gastroenterology & Hepatology
Article-Type of This Article
Observational Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Masood Muhammad Karim, Syedda Ayesha, Adeel Ur Rehman, Section of Gastroenterology, Department of Medicine, Aga Khan University, Karachi 74800, Sindh, Pakistan
Abdul Hadi Shahid, Muhammad Salman, Danaish Kumar, Medical College, Aga Khan University, Karachi 74800, Sindh, Pakistan
Zeeshan Uddin, Section of Histopathology, Department of Pathology and Laboratory Medicine, Aga Khan University, Karachi 74800, Sindh, Pakistan
Co-first authors: Masood Muhammad Karim and Syedda Ayesha.
Co-corresponding authors: Masood Muhammad Karim and Adeel Ur Rehman.
Author contributions: Karim MM and Rehman AU contribute equally to this study as co-corresponding authors. Karim MM and Ayesha S contribute equally to this study as co-first authors. Karim MM, Salman M, Ayesha S, and Kumar D contributed to structuring the manuscript; Karim MM, Ayesha S, and Shahid AH contributed to the project’s conception and design; Karim MM played a significant role in identifying the topic, providing valuable insights into the data; Ayesha S, Shahid AH, Salman M, and Kumar D contributed to collecting and cleaning data; Shahid AH conducted a literature search and wrote the manuscript; Salman M is interpreting the findings; Kumar D assisted in conducting a literature search; Uddin Z played a substantial role in interpreting the findings of fine-needle biopsy and histological slides, and proofreading the manuscript; Rehman AU supervised the project, provided a critical review, made final edits, and ensured its accuracy and clarity with his expertise.
Institutional review board statement: This study was reviewed and approved as an exemption by the Aga Khan University Hospital-Ethics Review Committee, with ethical code number No. 6874-25427.
Informed consent statement: Informed consent from patients was exempted in this study, as approved by the institutional ethical review committee, because the study did not involve any direct human interaction and utilized data obtained after each patient had provided written consent for treatment. Furthermore, all patient identifiers were removed during data entry, and patient confidentiality was fully maintained.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
STROBE statement: The authors have read the STROBE Statement-checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-checklist of items.
Data sharing statement: The datasets generated and/or analyzed during the current study are not publicly available owing to confidentiality and institutional ethical policies, but can be obtained from the corresponding author upon reasonable request.
Open Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Adeel Ur Rehman, Assistant Professor, Section of Gastroenterology, Department of Medicine, Aga Khan University, Karachi 74800, Sindh, Pakistan. adeelur.rehman@aku.edu
Received: September 2, 2025 Revised: October 2, 2025 Accepted: December 17, 2025 Published online: February 16, 2026 Processing time: 155 Days and 19.5 Hours
Abstract
BACKGROUND
The mediastinum comprises vital anatomical structures and can be affected by diverse benign and malignant pathologies. Imaging modalities such as computed tomography, magnetic resonance imaging, and positron emission tomography assist in initial evaluation, but histopathological confirmation remains the diagnostic gold standard. Endoscopic ultrasound (EUS) enables real-time, minimally invasive sampling of mediastinal lesions. EUS-guided fine needle biopsy (FNB) demonstrated high diagnostic accuracy with a favourable safety profile. Despite its established utility in high-income settings, data on the role of EUS in evaluating mediastinal pathologies in low- and middle-income countries, such as Pakistan, remain limited.
AIM
To assess the diagnostic yield of EUS-guided FNB in evaluating mediastinal pathologies and its role in guiding management in low- and middle-income countries.
METHODS
This single-centre retrospective study was conducted at Aga Khan University Hospital, Karachi, from June 2020 to June 2024, involving adults (≥ 18 years) who underwent EUS-guided FNB for mediastinal pathology. Franseen needles with capillary suction were used without rapid on-site evaluation, yielding adequate histopathological samples in all cases. Data included demographics, radiological, procedural, and histopathological details, treatment, and outcomes. Statistical analysis was performed using Statistical Package for the Social Sciences version 23, with receiver operating characteristic curve analysis used to assess the diagnostic performance of EUS-guided FNB.
RESULTS
A total of 77 patients (mean age: 54.4 ± 14.1 years; 55.8% male) underwent EUS-guided FNB for mediastinal pathology. Most (81.8%) were symptomatic, with lymphadenopathy or masses mostly detected on computed tomography (79.2%). Adequate tissue diagnosis was achieved in 94.8%; 41.6% were malignant. EUS-FNB showed 88.9% (95% confidence interval: 74.7-95.6) sensitivity, 100% (95% confidence interval: 91.4-100) specificity, and an area under the curve of 0.94. Malignant cases had significantly larger lymph nodes (2.3 ± 0.9 cm vs 1.57 ± 0.6 cm, P = 0.009) and masses (median 7.8 cm vs 4.2 cm, P = 0.001). Most EUS procedures (96.1%) were performed using a 22G needle. The mean number of needle passes was 2.1 ± 0.8. No post-procedure complications occurred. Common diagnoses included reactive lymphoid tissue (20.8%), tuberculosis (14.3%), and lymphoma (11.7%). Conservative therapy was the most common treatment (28.6%). Overall, 68.8% survived and 13.0% died.
CONCLUSION
EUS-guided FNB is a highly accurate and safe modality for diagnosing mediastinal pathologies, offering excellent diagnostic yield without procedure-related complications and effectively guiding management of both benign and malignant cases.
Core Tip: Endoscopic ultrasound-guided fine-needle biopsy (EUS-FNB) is a minimally invasive technique that enables accurate tissue sampling of mediastinal lesions. In this single-centre study from Pakistan, EUS-FNB demonstrated excellent diagnostic performance, with 94.8% diagnostic yield, 88.9% sensitivity, and 100% specificity, while maintaining a strong safety profile without procedure-related complications. The study highlights the value of EUS-FNB in guiding management of both benign and malignant mediastinal pathologies and addressing a critical evidence gap in low- and middle-income country settings, where data remain limited.
Citation: Karim MM, Ayesha S, Shahid AH, Salman M, Kumar D, Uddin Z, Rehman AU. Diagnostic accuracy and safety of endoscopic ultrasound-guided fine needle biopsy for evaluating mediastinal pathologies. World J Gastrointest Endosc 2026; 18(2): 113699
The mediastinum includes numerous essential anatomical structures and is centrally located within the thoracic cavity. It lies between the two pleural cavities and lungs on either side, bordered anteriorly by the sternum and posteriorly by the vertebral column, extending from the thoracic inlet to the diaphragm[1]. This region can be affected by a variety of pathological conditions, including both neoplastic and non-neoplastic lesions ranging from benign to malignant, and from primary to metastatic origins[2]. Many of these conditions manifest as mediastinal masses or lymph nodes, making the accurate diagnosis of such lesions a complex and demanding process for gastroenterologists, pulmonologists, radiologists, and pathologists alike[1]. Initial evaluation of mediastinal lesions (MLs) typically relies on imaging techniques such as computed tomography (CT), which helps characterize masses based on their composition, such as the presence of water, fat, or calcium. Magnetic resonance imaging provides additional detail by enhancing soft tissue contrast, allowing for better differentiation between benign and malignant MLs[3,4]. Although CT, magnetic resonance imaging, and positron emission tomography (PET) are effective in identifying primary tumors and detecting metastatic spread to mediastinal lymph nodes, their limitations in accurately staging these lesions are well recognized[5]. As a result, histological examination remains essential to confirm metastatic disease and to determine whether surgical resection is a viable treatment option[6]. Endoscopic ultrasound (EUS) has emerged as a valuable, minimally invasive technique for the diagnosis and staging of both benign and malignant conditions within the mediastinum. It also facilitates real-time, image-guided sampling of mediastinal lymph nodes and lesions, enhancing diagnostic accuracy[7,8]. EUS-guided fine needle aspiration or biopsy (EUS-FNA/FNB) offers several key advantages, including the ability to detect small lesions with ease, high sensitivity and specificity, and access to regions such as the posterior mediastinum and non-paratracheal lymph nodes. The procedure is relatively low-cost, straightforward to perform, and associated with a low risk of complications[9-12]. According to a multi-centre study, EUS-guided fine-needle biopsy (EUS-FNB) demonstrated a sensitivity of 90.2% and a specificity of 100% for diagnosing malignant tumors. The positive predictive value was 100%, the negative predictive value was 78.9%, and the overall diagnostic accuracy was 92.9%[13]. While EUS has established its role in mediastinal evaluation in high-income countries, data from low- and middle-income countries (LMICs) remain sparse, with very few studies focusing exclusively on mediastinal pathologies, particularly in our setting[14]. Its adoption in LMICs remains limited due to financial constraints, lack of expertise, and infrastructural challenges, thereby hindering widespread use and necessitating the critical need for context-specific evidence to guide clinical practice. In contrast to this, multiple studies from Western countries have consistently validated the accuracy and safety of EUS FNA/FNB. In 2023, a retrospective study of China showed that EUS-FNA/FNB is a reliable diagnostic tool for unidentified MLs, showing no notable complications. Similar findings were echoed in a multi-country study conducted in 2025[15,16]. These published studies from high-income countries emphasize the potential for expanding the use of EUS in LMICs as healthcare infrastructure and specialist training continue to develop. Therefore, our study aims to highlight the spectrum of mediastinal pathologies encountered, assess the diagnostic yield of EUS-guided tissue sampling, and explore its role in guiding further management. Our study seeks to contribute to the growing body of evidence supporting EUS as a valuable tool in the minimally invasive diagnosis of mediastinal conditions.
MATERIALS AND METHODS
A single-centre retrospective observational study was conducted in the gastroenterology section of the Department of Medicine at Aga Khan University Hospital (AKUH) in Karachi, Pakistan. The study included all inpatient and outpatient individuals who were detected with mediastinal pathology on radiological evaluation and subsequently underwent EUS-guided FNB of the lesion at the Endoscopy Suite for diagnostic purposes, between June 1, 2020 and June 30, 2024. After obtaining approval from the Institutional Ethical Review Committee, patient data were retrieved from the electronic medical records of adult patients who had undergone EUS-guided FNB for mediastinal lymphadenopathy or lesions. Patient records were accessed via the institutional health information management system. All patients aged 18 years and above, of either gender, with radiological or endoscopic evidence of mediastinal pathology who underwent EUS-FNB for further diagnostic evaluation were included in the study. Patients were excluded if they were below 18 years of age, with incomplete clinical data, had MLs that were inaccessible for EUS-guided sampling, or did not undergo EUS-FNB due to medical contraindications. The reviewed data included demographic and clinical information such as age, gender, symptom status, duration of symptoms, and previous history of malignancy. Radiological and endoscopic parameters encompassed the initial diagnostic modality used, the indication for EUS, and dimensions of the MLs or mediastinal lymph nodes. Procedural details included the type of FNB needle used (22-gauge or 25-gauge), the total number of needle passes performed, the overall duration of the EUS procedure, and any procedure-related complications. Additionally, in all our patients, the Franseen needle design with the capillary/slow pull suction method was used to obtain visible core samples for histopathology without rapid on-site evaluation (ROSE). All cases yielded adequate samples for histopathological diagnosis. Post-procedure, all patients were given oral antibiotics. Histopathological characteristics were reviewed to determine the obtained specimen, EUS diagnostic adequacy, the nature of the lesion (benign or malignant), and the definitive tissue diagnosis. Additionally, follow-up data, including treatment initiated based on the final diagnosis and the patient’s survival status until the last hospital encounter, were systematically documented. Descriptive statistical analysis was performed using the Statistical Package for the Social Sciences version 23. Categorical variables were presented as frequencies and percentages, while quantitative variables were expressed as mean ± SD for normally distributed data and as median with interquartile range for non-normally distributed data. The Shapiro-Wilk and Kolmogorov-Smirnov tests were used to assess the normality of numerical variables. The association between clinical, diagnostic, and procedural variables with the EUS-FNB tissue nature was evaluated using the χ2 test for categorical variables and the independent samples t-test for continuous variables. A P-value of < 0.05 was considered statistically significant. The diagnostic accuracy of EUS-FNB was evaluated, and the area under the curve obtained from the receiver operating characteristic curve analysis was used to assess the diagnostic performance of EUS-FNB in distinguishing between malignant and benign cases. This study was approved by the AKUH-Ethical Review Committee, with ethical code number No. 6874-25427. All patient identifiers were eliminated during data entry, and patient confidentiality was fully and completely respected.
RESULTS
A total of 77 patients who underwent EUS-guided FNB for mediastinal lymphadenopathy or MLs detected on radiological evaluation between June 2020 and June 2024 at our centre were included in this study. All patients were adults above 18 years of age, ranging from 22 years to 87 years. The mean age of the patients was 54.4 ± 14.1 years, with 55.8% (n = 43) being males and 44.2% (n = 34) females. Most patients, 77.9% (n = 60), had no previous history of malignancy. However, most patients, 81.8% (n = 63), were symptomatic at initial presentation. The mean duration of symptoms was 76.6 ± 69.2 days, ranging from 3 days to 305 days. The most common clinical presentations included abdominal pain in 26.0% (n = 20), shortness of breath in 19.5% (n = 15), jaundice in 15.6% (n = 12), cough in 14.3% (n = 11), weight loss in 11.7% (n = 9), vomiting in 10.4% (n = 8) and fever in 10.4% (n = 8) (Table 1). Initially, 79.2% (n = 61) of the patients underwent a CT scan, while 20.8% (n = 16) underwent a positron emission tomography scan. Based on initial diagnostic findings, the indications for EUS were mediastinal lymphadenopathy alone in 36.4% (n = 28); mediastinal and hilar lymphadenopathy with or without liver deposits or pulmonary nodules in 22.1% (n = 17); and mediastinal lymphadenopathy with or without liver deposits or gallbladder/pancreatic mass in 19.5% (n = 15). The mean lymph node size was 1.9 ± 0.8 cm (range: 0.7-3.6 cm), while the mean size of lesions or masses was 5.9 ± 2.4 cm (range: 1.3-9.4 cm). The mean EUS procedural duration was 28.1 ± 18.9 minutes. Additionally, a 22-G needle was used in 96.1% (n = 74) of cases. The mean number of needle passes was 2.1 ± 0.8, with 49.4% (n = 38) requiring two needle passes, followed by 26.0% (n = 20) requiring three needle passes, and 20.8% (n = 16) requiring one needle pass. No procedure-related complications were reported in any patient (Table 2). Tissue specimens obtained via EUS-FNB were sufficient to establish a diagnosis in 94.8% (n = 73) of cases. In most patients, the specimen was from mediastinal lymph nodes with or without esophageal, pancreatic, or splenic masses in 54.5% (n = 42), followed by MLs with or without esophageal or gastric lesions in 20.8% (n = 16). The EUS-FNB tissue nature was benign in 58.4% (n = 45) and malignant in 41.6% (n = 32). The most common definitive tissue diagnoses included reactive lymphoid tissue (20.8%, n = 16), tuberculosis (14.3%, n = 11), and lymphoma (11.7%, n = 9), followed by adenocarcinoma, squamous cell carcinoma, and sarcoidosis, each accounting for 10.4% (n = 8). Based on the final diagnosis, most patients received conservative therapy (28.6%, n = 22), followed by chemotherapy (18.2%, n = 14), anti-tuberculosis (TB) drugs (13.0%, n = 10), and palliative care (6.5%, n = 5). Overall, 68.8% (n = 53) of patients survived, 13.0% (n = 10) died, and 18.2% (n = 14) were either lost to follow-up or left against medical advice without receiving treatment (Table 3). In our study, 8.9% (n = 4) of patients with benign EUS-FNB tissue nature were later found to have a malignant definitive diagnosis, while none of the patients with malignant EUS-FNB tissue nature were later found to have benign disease. Thus, the sensitivity, specificity, positive predictive value, and negative predictive value of EUS-FNB were 88.9% [95% confidence interval (CI): 74.7-95.6], 100.0% (95%CI: 91.4-100), 100% (95%CI: 89.3-100), and 91.1% (95%CI: 79.3-96.5), respectively. Receiver operating characteristic curve analysis yielded an area under the curve of 0.94 (95%CI: 0.88-0.99) (Figure 1 and Table 4). In the malignant EUS-FNB group, patients had a higher mean age (56.7 ± 12.5 years) compared to the benign group (52.8 ± 15.1 years), though this was not statistically significant (P = 0.238). A history of previous malignancy was seen in 24.4% (n = 11) of benign cases vs 18.8% (n = 6) of malignant cases (P = 0.553). In the benign group, most patients, 84.4% (n = 38), were symptomatic, with a longer mean duration of symptoms (83.5 ± 75.5 days) compared to the malignant group, although this difference was not statistically significant. In the benign group, most patients, 44.4% (n = 20), had an indication for EUS as mediastinal lymphadenopathy only, compared to 25.0% (n = 8) in the malignant group (P = 0.254). The mean lymph node size was significantly larger in the malignant group (2.3 ± 0.9 cm) than in the benign group (1.57 ± 0.6 cm, P = 0.009). Similarly, the median lesion/mass size was significantly greater in the malignant group [7.8 cm (interquartile range = 3.4)] compared to the benign group [4.2 cm (interquartile range = 2.6), P = 0.001]. The EUS procedure duration was slightly longer in the benign group (29.6 ± 20.9 minutes) compared to the malignant group (25.8 ± 16.1 minutes), though this was not significant (P = 0.573). Tissue specimens obtained via EUS-FNB were insufficient to establish a diagnosis in 8.9% (n = 4) of benign cases, while all malignant cases yielded adequate specimens to conclude a diagnosis (P = 0.137). Among the benign group, conservative therapy was the most common treatment (46.7%, n = 21), followed by anti-TB drugs (17.8%, n = 8). In contrast, in the malignant group, most patients received chemotherapy (37.5%, n = 12), followed by palliative care and radiation therapy (each 9.4%, n = 3); this difference was statistically significant (P < 0.001). The benign group had a higher survival rate (77.8%, n = 35) and lower rates of loss to follow-up or left against medical advice (11.1%, n = 5), though these differences were not statistically significant (P = 0.104) (Table 5).
Figure 1 Receiver operating characteristic curve illustrating the diagnostic performance of the endoscopic ultrasound-guided fine needle biopsy.
The curve plots the true positive rate (sensitivity) against the false positive rate (1 - specificity). The area under the curve is 0.94 (95% confidence interval: 0.88-0.99), indicating excellent discriminatory ability of the procedure in distinguishing between malignant and benign cases. ROC: Receiver operating characteristic; AUC: Area under the curve.
Table 1 Demographics and clinical features of patients, n (%).
Table 5 Comparison of clinical, diagnostic, and procedural variables between patients with benign and malignant endoscopic ultrasound-guided fine needle biopsy tissue nature, n (%).
MLs arise from structures within the mediastinal compartments, located in the central thoracic cavity between the pleural cavities. It encompasses a heterogeneous spectrum of pathologies, ranging from reactive and infectious processes to benign and malignant neoplasms, and typically presents as mediastinal masses or enlarged lymph nodes[1,2]. Diagnostic evaluation is often challenging due to the diverse etiologies, limited accessibility for tissue sampling, and the close proximity of these lesions to numerous mediastinal structures[16]. However, the evolution of EUS and its technical advancements has significantly improved the diagnostic approach, offering a minimally invasive and reliable modality for the assessment and tissue acquisition of MLs[17]. Our retrospective study analyzed 77 patients who underwent EUS-guided FNB for mediastinal lymphadenopathy or lesions detected on radiological evaluation over a 4-year period at AKUH, a tertiary care hospital in Pakistan. In comparison, prior studies have reported variable sample sizes for EUS-guided tissue acquisition in mediastinal lymph node assessment, ranging from 72 cases in a single-centre Indian study over three years, to 107 cases in a Chinese retrospective study over four years, and up to 439 participants in a multicentre study across ten countries[2,15,17]. In this context, our study contributes a considerable number of EUS-guided FNB cases, aligning with previous studies, which may be attributed to our institution’s role as a tertiary care referral centre for the evaluation of MLs via EUS.
MLs encompass a wide histopathological and radiological spectrum across the person’s lifespan, though they are most frequently reported in middle-aged adults. This was also reflected in our cohort, where the mean age was in the 50s, consistent with previous studies[2]. Prior research has also demonstrated a male predominance among such patients, with Zhou et al[15] and Ardengh et al[18] reporting 71.0% and 72.5% male patients, respectively. In contrast, a multicentre study of 439 patients demonstrated a relatively lower proportion of males (58.1%), consistent with our findings of 55.8%[2]. Most of our patients, around 81.8% were symptomatic, with a relatively shorter mean duration of symptoms around 76.6 days, compared to a previous study by Dixit et al[19], which reported a mean duration of 110 days. Additionally, this study also reported shortness of breath as the second most common symptom, whereas in our study, the most frequently reported symptoms were abdominal pain, shortness of breath, jaundice, and cough. In contrast, other previous studies mentioned dyspnea, cough, and chest pain as the predominant clinical presentations[2,18]. Furthermore, the majority of our patients had Gastrointestinal lesions in addition to mediastinal lymphadenopathy, as detected on initial radiological imaging, predominantly via CT scan. Only a few patients, around 22.1% had hilar lymphadenopathy or pulmonary nodules in addition to mediastinal lymphadenopathy.
EUS-FNB is a minimally invasive technique that has emerged as a safe and valuable tool for the diagnosis of mediastinal space-occupying lesions, enabling distinction between benign and malignant pathologies[15,20]. In our study, EUS-FNB also proved to be a valuable modality for the evaluation of MLs, providing adequate tissue acquisition and sampling. Similarly, a previous randomized controlled trial demonstrated that EUS-FNB provided a comparable diagnostic yield while offering greater patient comfort than Endobronchial Ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in MLs evaluation[21]. In our LMIC setting, the ROSE facility was not readily accessible and was therefore not utilized. Previous studies have shown no significant difference in the diagnostic yield of EBUS-TBNA performed with or without ROSE, as cases deemed inadequate on ROSE were often still diagnosed successfully[22]. However, more recent data suggest improved diagnostic accuracy of MLs when ROSE is used[23]. Similarly, EUS-FNA has demonstrated higher diagnostic yield with ROSE, although its already high baseline accuracy makes further improvement difficult to establish[24]. In contrast, evidence from recent studies, including a systematic review, indicates that EUS-guided sampling with MOSE provides diagnostic yield and accuracy comparable to ROSE and has been reported as a reliable alternative in patients undergoing EUS-FNB[25-27]. Consistent with these findings, our study showed an overall diagnostic adequacy of 94.8%, including 100% for malignant lesions, comparable to the 93.0% reported in patients with MLs who underwent EUS-FNA with ROSE[17].
Importantly, no procedure-related complications were observed in our cohort, in contrast to previous studies that have reported complication rates ranging from 0.1% to 2%[2]. Additionally, our study demonstrated high sensitivity, specificity, positive predictive value, and negative predictive value for EUS-FNB, findings consistent with those reported in previous multi-centre studies[2,28,29]. In contrast, a recent study demonstrated lower sensitivity of EBUS-TBNA in diagnosing hilar and mediastinal pathologies, largely attributable to non-sampling of accessible lymph nodes[30]. In the literature, needle-related aspects, particularly needle size and gauge size, have been identified as important determinants of diagnostic yield in EUS-FNB[2]. Several studies have associated smaller diameter needles with improved cytological yield, with one study reporting a diagnostic accuracy of 81% using a 22G needle[31]. In our cohort, the majority of our patients, around 96.1% utilized a 22G needle. Moreover, prior studies have shown that an increased number of needle passes enhances diagnostic performance by improving the adequacy of retrieved tissue specimens[2,32,33]. In our study, the average number of needle passes was around 2, consistent with the findings of Fernández-Esparrach et al[34], while the maximum number of needle passes was 5. By comparison, another study reported up to six passes during the MLs’ evaluation[2].
Most of our MLs, around 58.4% were benign in consistent with findings of Mukundan et al[17], who reported granulomatous inflammation as the most common diagnosis. In contrast, a study by Zhou et al[15] reported 65.4% cases of malignancy, with lung cancer being the most frequent diagnosis. In our study, patients with a malignant diagnosis were found to be of a higher age as compared to the benign diagnosis group, though the difference was not statistically significant. Previous studies have shown that elderly and male patients were more likely to have malignant disease[2,35]. Additionally, patients with a benign diagnosis were more often symptomatic and had a longer duration of symptoms compared to the malignant group in our study. In contrast, B et al[36] showed that patients in the malignant group were more symptomatic with longer duration of symptoms. Additionally, the size of lymph nodes and lesions was significantly larger for malignant diseases as compared to benign ones, as seen in previous studies too. Junare et al[20] reported an average size of lymph node of 3.5 cm in cases of tubercular lymphadenitis, whereas in our benign group, the average size of lymph node was found to be 1.9 cm. Furthermore, a greater proportion of patients in the malignant group required 3 number of needles passes as compared to benign group. These corresponded with a diagnostic adequacy of 100% for malignant lesions vs 91.1% for benign lesions, as also reported in a previous study. However, it could also be attributed to tumor heterogeneity, or histological limitations such as interpretative challenges or sampling error[37,38]. Importantly, none of the malignant EUS-FNB diagnoses were later revised as benign, highlighting the reliability of malignant results. Of the four cases initially misdiagnosed as benign on EUS-FNB, one was later diagnosed as metastatic carcinoma, while the remaining three were confirmed as lymphoma following lymph node resection.
In our study, reactive lymphoid tissue was the most common benign diagnosis, followed by TB and sarcoidosis. Evison et al[39] reported a reactive lymphadenopathy diagnosis in 48% of mediastinal and hilar lymphadenopathy cases. Overall, TB was the second most common reported diagnosis in our study, likely due to its endemicity in our region, findings consistent with the study by Junare et al[20]. Likewise, Zhou et al[15] showed TB followed by sarcoidosis as the most frequent benign lesions. Whereas, B et al[36] reported lymphoma in 22.7% of all cases as the most common one. Consistent with this, lymphoma emerged as the top definitive malignancy in 11.7% of total cases. Junare et al[20] demonstrated 100% sensitivity and specificity of EUS-FNA/FNB in diagnosing lymphoma, consistent with our findings of 100% diagnostic adequacy among patients in the malignant group. Additionally, B et al[36] identified thymoma in 13.6% of cases as the third most common diagnosis, whereas our study only showed 2.6% thymoma of total cases.
Importantly, these definitive diagnoses established via EUS-FNB directly guided patient management, thereby avoiding unnecessary invasive or traumatic diagnostic procedures. A previous study reported that 18 patients were spared from invasive procedures due to EUS, and even some patients were referred for chemotherapy due to small-cell lung cancer established via EUS-FNA[40]. Similarly, Del Vecchio Blanco et al[41] highlighted the role of EUS findings in guiding chemotherapy in 76%, avoiding surgery in 16%, and adopting conservative therapy in 2%. In our study, most of our patients, around 28.6% underwent conservative therapy, 13% were initiated on anti-TB treatment, and 18.2% received chemotherapy, while surgical resection was performed in only 3.9% of cases where indicated.
It is also important to note that EUS utilization in LMICs is constrained by several practical barriers, including high equipment and maintenance costs, limited availability of trained endosonographers, inadequate infrastructure, and the lack of multidisciplinary support[42,43]. Moreover, healthcare resources in these settings are often directed toward more pressing public health priorities, leaving advanced diagnostic modalities like EUS underfunded and underutilized. Consequently, clinical practice guidelines from high-income countries, which recommend routine integration of EUS for diagnosis, staging, and tissue acquisition, are often impractical in LMICs, where patients may face delays due to referral to distant, overburdened tertiary centres[44,45]. Context-specific guidelines are therefore required, balancing proven clinical benefits with financial feasibility, workforce training limitations, and local healthcare priorities. In this context, our study contributes to the emerging LMIC data by providing context-specific evidence while emphasizing the need for larger prospective studies to support the development of resource-sensitive, evidence-based guidelines tailored to local realities.
Additionally, this study has certain limitations that should be acknowledged. Being a single-centre retrospective study, the analysis relied on existing records, which may have led to variability in documentation detail. The relatively small sample size, restricted to patients presenting to our centre, introduces selection bias and limits the generalizability of the findings. In addition, subgroup analyses included multiple comparisons without adjustment due to the small sample size, as strict corrections would have reduced statistical power. Therefore, reported p-values should be interpreted with caution, though the findings remain clinically meaningful. Furthermore, the lack of long-term follow-up with detailed outcomes and cost-effectiveness analysis, inherent to the study design, limited the ability to fully evaluate its diagnostic and therapeutic impact, particularly in LMIC settings. Consequently, future longitudinal multi-centre studies with larger cohorts are warranted to more definitively establish the diagnostic accuracy and management influence of EUS in mediastinal pathologies evaluation.
CONCLUSION
EUS-guided FNB proved to be a safe and effective diagnostic modality for evaluating MLs in our cohort, with high diagnostic adequacy, particularly for malignant cases, and no procedure-related complications. Most lesions were benign, reflecting regional disease patterns such as tuberculosis, while lymphoma emerged as the most frequent malignant diagnosis. Needle-related factors, including appropriate needle selection and an adequate number of passes, ensured reliable tissue acquisition. Importantly, EUS-FNB findings directly influenced patient management, allowing many to be treated conservatively or with medical therapy, thereby avoiding unnecessary invasive procedures. These results, consistent with prior regional and international studies, reinforce the role of EUS-FNB as a first-line minimally invasive tool in the diagnostic and therapeutic pathway of MLs, particularly in tertiary care referral centres.
Footnotes
Provenance and peer review: Invited article; Externally peer reviewed.
Peer-review model: Single blind
Corresponding Author's Membership in Professional Societies: Member World Endoscopy organisation WEO; Member American College of gastroenterology ACG; Member American Society of Gastrointestinal Endoscopy ASGE; and Member Gastroenterological Society of Singapore GESS.
Specialty type: Gastroenterology and hepatology
Country of origin: Pakistan
Peer-review report’s classification
Scientific Quality: Grade B, Grade C
Novelty: Grade C, Grade C
Creativity or Innovation: Grade B, Grade C
Scientific Significance: Grade B, Grade C
P-Reviewer: Rath S, MD, Senior Researcher, India S-Editor: Bai SR L-Editor: A P-Editor: Zheng XM
Okasha H, Gadour E, Tayyab GUN, Seicean A, Hicham E, Emad E, Alamri T, Tehami N, Faraj HI, Alzamzamy A, Hussein H, Tag-Adeen M, Elbasiony M, Al-Lehibi A, Saleh S, Houmani Z, Abdelghani M, Ragab K, Ismaili Z, Drir O, Ghoneem E, Abdallah O, Rebiai S, Pausawasdi N, Rouibaa F, Borahma M, Atalla H, Farouk M, Altonbary A. Comprehensive analysis of endoscopic ultrasound (EUS) in mediastinal lesions: patient characteristics, diagnostic outcomes, safety, and efficacy: a multi-country study.Egypt J Intern Med. 2025;37:25.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 1][Reference Citation Analysis (0)]
Ebner R, Sheikh GT, Brendel M, Ricke J, Cyran CC. ESR Essentials: staging and restaging with FDG-PET/CT in oncology-practice recommendations by the European Society for Hybrid, Molecular and Translational Imaging.Eur Radiol. 2025;35:1894-1902.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 6][Reference Citation Analysis (0)]
Iglesias-Garcia J, Poley JW, Larghi A, Giovannini M, Petrone MC, Abdulkader I, Monges G, Costamagna G, Arcidiacono P, Biermann K, Rindi G, Bories E, Dogloni C, Bruno M, Dominguez-Muñoz JE. Feasibility and yield of a new EUS histology needle: results from a multicenter, pooled, cohort study.Gastrointest Endosc. 2011;73:1189-1196.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 219][Cited by in RCA: 233][Article Influence: 15.5][Reference Citation Analysis (0)]
Mushtaq HI, Shams F, Rasool S, Hassan GU, Jabbar S, Javed F, Rasheed S, Dlishad A, Tayyab GUN. Experience of Endoscopic Ultrasound Guided Fine Needle Aspiration and Fine Needle Biopsy: Data from Tertiary Care Hospital in Pakistan.Pak J Health Sci. 2024;5:31-35.
[PubMed] [DOI] [Full Text]
Mukundan S, Venkatakrishnan L, Vishnu Abishek R. Unveiling mediastinal pathology: role of EUS guided fine needle aspiration in diagnosing mediastinal lesions.Int J Res Med Sci. 2018;6:2952.
[PubMed] [DOI] [Full Text]
Magdy O, Abdeldayem A, Elmaraghi A, Ali M, Hafez F. The value of rapid on-site evaluation during conventional and endobronchial ultrasound needle aspiration in the diagnosis of mediastinal lymphadenopathy and lung cancer.Egypt J Bronchol. 2024;18:70.
[PubMed] [DOI] [Full Text]
Sundaram S, Chhanchure U, Patil P, Seth V, Mahajan A, Bal M, Kaushal RK, Ramadwar M, Prabhudesai N, Bhandare M, Shrikhande SV, Mehta S. Rapid on-site evaluation (ROSE) versus macroscopic on-site evaluation (MOSE) for endoscopic ultrasound-guided sampling of solid pancreatic lesions: a paired comparative analysis using newer-generation fine needle biopsy needles.Ann Gastroenterol. 2023;36:340-346.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 7][Cited by in RCA: 11][Article Influence: 3.7][Reference Citation Analysis (0)]
Gadour E, Miutescu B, Al Ghamdi S, Burciu C, Shaaban H, Vuletici D, Almuhaidb A, Ratiu I, Aljahdli E, Okasha H. Diagnostic accuracy and sensitivity of the rapid on-site evaluation (ROSE) versus macroscopic on-site evaluation (MOSE) in endoscopic ultrasound (EUS)-guided sampling: a systematic review.Frontline Gastroenterol. 2025;16:489-494.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 1][Reference Citation Analysis (0)]
Facciorusso A, Arvanitakis M, Crinò SF, Fabbri C, Fornelli A, Leeds J, Archibugi L, Carrara S, Dhar J, Gkolfakis P, Haugk B, Iglesias Garcia J, Napoleon B, Papanikolaou IS, Seicean A, Stassen PMC, Vilmann P, Tham TC, Fuccio L. Endoscopic ultrasound-guided tissue sampling: European Society of Gastrointestinal Endoscopy (ESGE) Technical and Technology Review.Endoscopy. 2025;57:390-418.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1][Cited by in RCA: 20][Article Influence: 20.0][Reference Citation Analysis (1)]
