The 2024 global tuberculosis report identifies China as a high-burden country for tuberculosis (TB) and multidrug-resistant TB (MDR-TB), accounting for 7.3% of global rifampicin-resistant TB (RR-TB) cases (World Health Organization, 2024). Cervical (neck) tuberculous lymphadenitis constitutes 30-40% of extrapulmonary tuberculosis (EPTB) cases (Ganchua et al., 2020; Li et al., 2023; Zhao et al., 2024) and is characterized by localized chronic infection, granulomatous inflammation, and caseous necrosis (Flyger et al., 2020). Despite its clinical prevalence, diagnostic challenges persist, particularly in general hospitals where the proportion of extrapulmonary tuberculosis diagnoses remains relatively low. In contrast, specialized tuberculosis centers report higher diagnostic rates due to enhanced resources and professional capabilities (Ohene et al., 2019; Fang et al., 2022). This disparity arises from insufficient clinical vigilance, limited access to molecular diagnostics, and the histopathological mimicry of granulomatous lesions by non-tuberculous mycobacterial infections, sarcoidosis, and fungal diseases (Ohene et al., 2019; Yenilmez et al., 2021; Jain et al., 2024). Consequently, the absence of microbiological confirmation in certain patients undergoing surgical resection of cervical lymph nodes results in delayed diagnosis and suboptimal management of drug-resistant tuberculosis (DR-TB) (Lekhbal et al., 2020; Soriano et al., 2024). Improving etiological detection rates and the early identification of drug resistance remains a critical challenge (Ben Ayed et al., 2019; Li et al., 2023). This study investigates how pathological types and sampling methods influence etiological positivity rates in cervical tuberculous lymphadenitis (CTL) and proposes strategies to enhance diagnostic accuracy.

Granulomatous inflammation represents a non-specific pathological feature observed not only in tuberculous lymphadenitis but also in non-tuberculous mycobacterial infections, sarcoidosis, cat-scratch disease, and certain fungal infections (Öztomurcuk et al., 2022). Consequently, the presence of granulomatous inflammation alone cannot establish a definitive diagnosis of tuberculous lymphadenitis. The pathological progression of cervical tuberculous lymphadenitis demonstrates significant complexity: initial caseous necrosis may evolve into liquefactive abscess formation (Srinivas and Nair, 2019), and capsular rupture of necrotic components can lead to soft tissue infiltration, while calcification typically occurs during the healing phase (Yang and Du, 2019). Nevertheless, atypical manifestations such as coagulative necrosis or non-necrotizing lymphadenitis may emerge in scenarios involving robust host immunity, early disease stages, or low bacterial burden (Srinivas and Nair, 2019; Yang and Du, 2019). The diagnosis of cervical tuberculous lymphadenitis presents multiple challenges: firstly, the non-specificity of pathological characteristics (Shah et al., 2017); secondly, the persistent inability to achieve etiological confirmation even in partially resected lymph node specimens (Kim et al., 2025). These limitations not only compromise treatment adherence but may also delay the diagnosis and management of drug-resistant tuberculosis (Wulandari et al., 2024). Therefore, enhancing diagnostic yield in extrapulmonary tuberculosis specimens remains imperative. Through systematic analysis of 198 surgical cases with cervical tuberculous lymphadenitis, this study pioneers the identification of pathological subtypes and sampling methodologies as critical determinants influencing etiological detection outcomes, thereby proposing novel strategies to address diagnostic dilemmas in extrapulmonary tuberculosis.

Current clinical diagnostic techniques for TB exhibit varying sensitivities and clinical utilities. AFB staining demonstrates low positivity rates in specimens obtained from cervical tuberculous lymphadenitis. As highlighted in the study by Mukhida et al., extrapulmonary TB specimens such as pleural fluid show Ziehl-Neelsen staining positivity rates of merely 0-40%, which aligns with our findings (Mukhida et al., 2025). Notably, when AFB-positive results coincide with TB-DNA negativity, this combination may indicate potential non-tuberculous mycobacterial infections (Kunkel et al., 2016). Published evidence reveals that SAT-TB achieves higher detection rates than AFB in morning sputum specimens from pulmonary TB cases, with positive results suggesting mycobacterial viability and demonstrating strong concordance with clinically confirmed TB diagnoses (Yan et al., 2018). Our findings indicate that SAT-TB exhibits lower sensitivity compared to the Xpert assay, which precludes its use as a first-line diagnostic tool for cervical tuberculous lymphadenitis. However, its high negative predictive value (91.59%) suggests potential utility as a predictive biomarker for MTB culture-negative status in lymph node specimens. Meanwhile, TB-DNA targeting species-specific nucleic acid sequences, exhibits superior specificity alongside favorable cost-effectiveness and accessibility in routine clinical practice (Kim et al., 2023). The Xpert MTB/RIF assay demonstrates high sensitivity and specificity, while also providing evidence for rifampicin resistance. It has been widely utilized in diagnosing pulmonary tuberculosis and osteoarticular tuberculosis (Sun et al., 2019). In our study, the GeneXpert MTB/RIF assay achieved an etiological positivity rate of 90.36% (P < 0.001 vs. other methods). The elevated positive detection rate for etiology is likely attributable to our “liquefactive necrosis-priority” sampling principle. Previous studies have demonstrated that the Xpert MTB/RIF assay exhibits superior diagnostic capability for rifampicin resistance detection compared to the MTB-culture assay (Mukhida et al., 2022). Our study revealed consistent results between Xpert rifampicin resistance determinations and MTB-culture drug susceptibility testing. Therefore, the Xpert assay is recommended as the preferred method for both diagnosing cervical lymph node tuberculosis and initial screening for rifampicin resistance. Although the positive detection rate of MTB culture was only 16.67%, but the drug susceptibility testing results obtained from it play a pivotal role in guiding the formulation of subsequent treatment regimens.

Among 179 specimens subjected to five etiological detection methods, those exhibiting liquefactive necrosis demonstrated significantly higher etiological detection rates via AFB, TB-DNA, SAT-TB, and Xpert methods compared to specimens with only solid alterations. This indicates a strong correlation between microbiological positivity rates and pathological alterations. In abscess-type pathological subtypes, AFB achieved a positivity rate exceeding 20%, while TB-DNA surpassed 85%. Moreover, Xpert exhibited a 100% detection rate in incision/drainage specimens from abscesses. This enhanced sensitivity likely arises from the abundant necrotic material characteristic of abscess formation. Furthermore, specimens with coexisting coagulative necrosis and liquefactive necrosis showed a lower TB-DNA positivity rate (66.67%) compared to pure liquefactive necrosis specimens (81.53%) (P = 0.032). Collectively, these findings suggest that a greater proportion of liquefactive necrosis correlates with higher microbiological yield and MTB burden. These results align with reports indicating enhanced MTB survival in necrotic neutrophils compared to intact cells (Liang et al., 2020). These findings also indicates that neutrophil hydrolases break down necrotic matrices, releasing MTB antigens or DNA, thereby enhancing the sensitivity of molecular tests such as PCR. Additionally, MTB inhibits neutrophil apoptosis, prolonging neutrophil survival and further promoting bacterial proliferation (Andrews et al., 2024). Compared to other fluid necrosis specimens, calcified liquid necrosis specimens exhibit lower pathogen detection rates across various testing methods ( Supplementary table 1 ). This may be related to the calcified encapsulation of necrotic material, which prevents pathogens from being fully released by conventional digestive solutions (Sharma et al., 2022). The underlying mechanisms and this conclusion require further in-depth investigation.

Therefore, we recommend prioritizing specimens exhibiting liquefactive necrosis from cervical lymphadenitis for laboratory testing. To optimize the acquisition of liquefactive necrotic specimens, the following recommendations are proposed. Firstly, clinicians should perform color contrast-enhanced ultrasound punctures to accurately access liquefied necrotic zones, guided by real-time hypoechoic imaging findings (Zhang et al., 2023). The Xpert system demonstrated a diagnostic yield of 93.3% (237/254) for etiological identification in specimens obtained via contrast-enhanced ultrasound-guided biopsy (Sun et al., 2021). Secondly, surgical teams must prioritize the collection of caseous necrosis under direct visualization. Furthermore, for both fresh specimens and formalin-fixed paraffin-embedded tissue samples, pathologists should collaborate closely with laboratory technicians to maximize the selection of necrotic components. These findings provide theoretical support for optimizing specimen collection strategies and enhancing pathogen detection rates.

Our findings indicate that specimens obtained through incision and drainage exhibit higher rates of TB-DNA positivity and Xpert assay detection compared to surgical specimens. These differences may be attributed to the enhanced accessibility of deep necrotic tissue during incision and drainage procedures. Furthermore, the combination of specimens from incision and drainage with surgical resection specimens, or the integration of puncture specimens with surgical resection specimens, results in higher overall pathogen detection rates than employing a single sampling method. This is particularly critical for MTB-culture diagnostic approaches, which inherently have low etiological positivity rates. Therefore, clinicians are advised to adopt a multimodal sampling strategy (e.g., combining incision and drainage with surgical resection sampling) for MTB-culture testing to enhance detection sensitivity for pathogen.

This study highlights the critical role of pathological features and sampling techniques in optimizing the etiological detection of cervical lymph node tuberculosis. The Xpert assay demonstrated superior sensitivity (90.36%), significantly outperforming four other diagnostic techniques, particularly in specimens with liquefactive necrosis (caseous necrosis, abscess, or soft tissue involvement), where its positivity rate exceeded 95%. Pathological analysis revealed that specimens with liquefactive necrosis exhibited markedly higher detection rates across AFB, TB-DNA, SAT-TB, and Xpert assay compared to those with solid alterations, emphasizing the necessity of prioritizing necrotic tissue during sampling. Furthermore, combining incision/drainage specimens with surgical resection specimens improved MTB culture positivity rates, addressing its inherent low sensitivity. The Xpert assay also reliably detected rifampicin resistance, aligning with culture-based drug susceptibility results. These findings advocate for a stratified diagnostic approach, prioritizing the Xpert assay for specimens exhibiting liquefactive necrosis to maximize sensitivity. Additionally, employing a combination of multiple sampling techniques—such as surgical resection, incision and drainage, or needle biopsy—can enhance the positive rate of etiology. Such strategies are crucial in countries burdened by extrapulmonary tuberculosis (TB) as they expedite diagnosis, guide therapy, and mitigate the risks of drug resistance. Future multicenter studies should aim to validate these findings and explore the relationship between the extent of necrosis and resistance patterns to refine diagnostic frameworks.