Elevated nil values in TB-IGRA: a potential diagnostic clue for pediatric histiocytic necrotizing lymphadenitis

Abstract

Background

Histiocytic necrotizing lymphadenitis (HNL) is a benign condition in children that is often misdiagnosed due to its clinical overlap with tuberculosis, rheumatic diseases, and lymphadenitis. The QuantiFERON-TB Gold In-Tube assay (TB-IGRA) is routinely used to rule out tuberculosis, but its Nil value (TB-Nil), which reflects background IFN-γ release, remains underutilized. Many studies suggest that elevated IFN-γ levels may aid in the differential diagnosis of HNL. In this study, we investigate the role of TB-Nil in diagnosing HNL.

Methods

To investigate the relationship between TB-Nil values and the diagnosis of HNL, we retrospectively enrolled 48 patients who were diagnosed with HNL at the Women and Children’s Medical Center, Guangzhou Medical University, between January 2019 and December 2024 and who had undergone TB-IGRA testing. Clinical data were extracted from the electronic medical record system. TB-Nil values were compared between HNL and other common causes of lymphadenitis, including systemic juvenile idiopathic arthritis (sJIA), Kawasaki disease (KD), and bacterial lymphadenitis (BL). Receiver operating characteristic (ROC) curve analysis was performed to evaluate diagnostic performance.

Results

Among the 48 enrolled patients with suspected HNL, 37 were confirmed by pathological diagnosis and the remaining 11 were diagnosed based on clinical criteria. The median Nil level was significantly elevated in HNL patients [0.82 IU/mL, interquartile range (IQR): 0.36–1.18] compared with patients with systemic juvenile idiopathic arthritis (sJIA, median = 0.07 IU/mL, P < 0.0001), Kawasaki disease (KD, median = 0.09 IU/mL, P < 0.0001) and Bacterial Lymphadenitis (BL, median = 0.09 IU/mL, P < 0.0001). Receiver operating characteristic (ROC) curve analysis revealed favorable diagnostic performance of the Nil index for identifying HNL, with an AUC of 0.8781 (95%CI: 0.8138–0.9424). The optimal cutoff value of TB-Nil was determined as 0.19 IU/mL for distinguishing HNL from the three control diseases (sJIA, KD and BL), corresponding to a sensitivity of 78.38% and a specificity of 91.67%.

Conclusion

The routinely available yet often overlooked Nil value from the TB-IGRA is significantly elevated in pediatric HNL and demonstrates favorable diagnostic accuracy. It may serve as a practical, non-invasive adjunctive biomarker for the early differential diagnosis of HNL from other inflammatory diseases. Prospective multi-center studies are warranted to validate these findings.

Background

Histiocytic necrotizing lymphadenitis (HNL), also known as Kikuchi-Fujimoto disease (KFD), is a benign and self-limiting disorder characterized by cervical lymphadenopathy and prolonged fever [1]. Although its etiology remains elusive, it is currently regarded as an immune-mediated condition, potentially triggered by viral or other infectious agents in genetically susceptible individuals [2]. The clinical presentation of HNL closely mimics that of tuberculosis (TB), viral lymphadenitis, various rheumatic immune diseases (such as systemic lupus erythematosus [SLE], systemic juvenile idiopathic arthritis [sJIA], Kawasaki disease [KD], and bacterial lymphadenitis[BL]), as well as malignant lymphoma (ML) [3]. Because laboratory findings are non-specific, HNL is frequently misdiagnosed or subject to delayed diagnosis. Consequently, patients often undergo extensive and prolonged diagnostic workups before a lymph node biopsy provides a definitive answer [4].

The diagnostic challenge is particularly pronounced in pediatrics [5]. Recently, serum cytokine profiling has emerged as a potential non-invasive diagnostic aid [6]. Studies have suggested that in children with fever of unknown origin and lymphadenopathy, serum IFN-α levels are elevated in HNL patients and positively correlate with disease activity [7]. Elevated serum IFN-γ, IFN-γ/IL-6 ratio, and serum IL-6 levels have also been observed in HNL [8]. However, these findings require validation in larger, multi-center cohorts.

The QuantiFERON-TB Gold In-Tube assay, a whole-blood interferon-gamma release assay (IGRA), is routinely employed worldwide to the diagnose of TB infection [9]. It measures IFN-γ release from memory T-cells after stimulation with TB-specific antigens [10]. The test comprises three components: the TB antigen tube (Ag), the mitogen positive control tube (MT), and the nil tube (NL) [11]. Its sensitivity for diagnosing tuberculosis infection ranges from 91% to 94%, with a corresponding specificity of 97% to 98% [12], indicating that this assay has excellent diagnostic performance for TB infection. However, the elevated Nil values are often dismissed as “background noise” or merely “indeterminate”, may reflect the patient’s underlying state of immune activation [11, 13].

In this study, we aimed to reinterpret these readily available yet underutilized data. By comparing Nil values in children with HNL against those with other common causes of lymphadenopathy, we sought to determine whether this simple, routinely collected parameter could serve as an early, non-invasive clue for diagnosing pediatric HNL.

Patients and methods

Study population and design

This study included 48 patients with HNL, as well as age-matched controls consisting of 31 patients with sJIA, 21 with KD, and 22 with BL. All patients were diagnosed at the Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, between January 2019 and December 2024, and all had undergone TB-IGRA testing. The diagnosis of HNL was based on either histopathological results from lymph node biopsy [14] (biopsy-proven HNL, n = 37) or, in the absence of biopsy, a decision by a pediatric hematologist or rheumatologist based on typical clinical, laboratory, and imaging features (clinical HNL, n = 11). These features included fever lasting > 5 days, painful cervical lymphadenopathy, lack of response to antibiotics for at least 72 h, leukocytopenia, and evidence of multiple enlarged lymph nodes on ultrasound [7]. Patients were excluded if they had positive TB-IGRA or purified protein derivative (PPD) skin test results, a history of tuberculosis exposure, fever requiring anti-tuberculosis therapy for resolution, confirmed concomitant pathogen infection via blood culture, throat swab or NGS, or concomitant rheumatic disorders including SLE.

sJIA was diagnosed following the 2001 International League Against Rheumatism criteria [15]. KD was diagnosed based on established diagnostic criteria [16]. The diagnostic criteria for BL included fever, lymphadenopathy, elevated CRP, leukocytosis with neutrophilic predominance, and ultrasound findings suggestive of suppurative changes. This study was approved by the Institutional Review Board of Guangzhou Women and Children’s Medical Center, Guangzhou Medical University. Due to the retrospective nature of the study, informed consent from patients or their guardians was waived.

TB-IGRA measurement

TB-IGRA examination was performed at the time of the patient’s first visit, before the initiation of any treatment with corticosteroids or immunosuppressants. The Nil values from the TB-IGRA were recorded. Demographic information for each patient, including gender, age, symptoms, signs, and clinical diagnosis, was also recorded.

Data collection

Clinical data were collected from inpatient and outpatient medical records. Data collected included clinical manifestations and laboratory findings. Clinical and laboratory data were collected at the onset of HNL. Clinical manifestations are listed in Table 1. Laboratory variables included total leukocyte, neutrophil, lymphocyte, and platelet counts; erythrocyte sedimentation rate (ESR); hemoglobin levels; C-reactive protein (CRP); aminotransferases; ferritin; fibrinogen; lactate dehydrogenase (LDH); triglycerides; IL-6; and IFN-γ.

Table 1.

Laboratory features of the HNL patients

	HNL subjects (N=48)	HNL-Definitive (N=37)	HNL-Clinical (N=11)	P values (HNL-Definitive vs. HNL-Clinical )
Male (%)	31,64.6%	27, 72.9%	4, 36.3%	0.012
Age, years old	10(7-12)	10(7-12)	10(7.5-12)	0.3771
WBC, 109/L	3.85(2.8-4.9)	4.4(2.9-5.8)	3.2(2.4-4.4)	0.0993
Neutrophils,109/L	1.66(1.2-2.6)	1.8(1.3-2.8)	1.2(0.8-2.4)	0.0872
Lymphocytes,109/L	1.5(1.1-1.9)	1.5(1.1-1.9)	1.3(1.1-1.8)	0.9205
Hemoglobin, g/dL	108(98-115)	108(97-116)	111(103-116)	0.5088
Platelets, 109/L	227(181-303)	232(186-294)	194(195(176-325)	0.6327
CRP, mg/dL	5.8(2.4-14.1)	5.8(2.5-17.6)	3.7(2.4-9.3)	0.5465
AST, IU/L	20(12.0-31.7)	21(11-32)	15(12-27)	0.5167
Triglycerides, mg/dL	1.1(0.9-1.4)	1.1(0.9-1.4)	0.91(0.75-1.54)	0.3212
ESR (mm/h)	53(31-67)	51(32-70)	56(20-66)	0.8575
LDH (IU/mL)	386(293-478)	381(293-492)	391(305-416)	0.5219
TB	4.7(3.9-6.3)	4.5(3.9-6.8)	4.8(3.6-5.9)	0.9556
IgG(g/L)	13.1(12.0-15.3)	13.3(11.9-15.3)	13(11.9-17.8)	0.9738
IgA(g/L)	2.4(1.5-2.9)	2.5(1.6-3.1)	1.8(1.3-2.5)	0.2106
IgM(g/L)	1.2(0.8-1.9)	1.2(0.8-2.0)	1.17(0.9-1.5)	0.6828
C3(g/L)	1.2(1.0-1.3)	1.2(1.1-1.3)	1.3(1.0-1.4)	0.5618
C4(g/L)	0.34(0.27-0.41)	0.34(0.27-0.41)	0.34(0.29-0.44)	0.4805
INF-γ(pg/ml)	14.35(6.9-22.9)	14.9(11.8-36.8)	14.65(6.8-75.5)	0.8417
Ferritin(ng/mL)	281(145-488)	277(145-525)	291(110-476)	0.714
Fibrinogen, mg/dL	4.0(3.4-4.4)	3.99(3.47-4.50)	3.8(3.3-4.2)	0.4017
TB-Nil(IU/mL)	0.82(0.37-1.19)	0.79(0.33-1.21)	0.98(0.48-1.16)	0.5653

WBC, white blood cell counts; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; LDH, lactate dehydrogenase; AST, aspartate aminotransferase; TB, total bilirubin; IgG, immunoglobulin G; IgA, immunoglobulin A; IgM, immunoglobulin M; C3,complement3; C4, complement 4; IL-6, interleukin-6; IFN-γ, interferon-γ

Statistical analysis

Statistical analysis was performed using GraphPad Prism 9.5.1 (GraphPad, San Diego, CA, USA). Continuous variables were presented as medians with interquartile ranges (IQR), and categorical variables were presented as frequencies and percentages. Groups were compared using Fisher’s exact test, and the Mann-Whitney U test. Spearman correlation was applied to evaluate correlations between TB-Nil and IFN-γ, P-values < 0.05 were considered statistically significant. The optimal cutoff values for receiver operating characteristic (ROC) curve analysis were determined by Youden’s index, which maximizes (sensitivity + specificity – 1).

Results

Clinical characteristics of HNL patients

As shown in Table 1, the 48 enrolled HNL children included 37 pathologically confirmed and 11 clinically diagnosed cases. Male predominance was noted in the overall cohort (64.6%); Median age was 10 years for both subgroups, without significant age difference (P = 0.3771).All 48 enrolled patients underwent TB‑IGRA testing and tested negative.

No significant between-group differences in clinical and laboratory parameters were observed when comparing patients with histopathologically confirmed HNL versus clinically diagnosed HNL via the Mann–Whitney U test. Common laboratory abnormalities comprised elevated erythrocyte sedimentation rate (ESR, median: 53 mm/h, range: 31–67 mm/h), mild leukocytopenia (median: 3.8*10⁹/L, range: 2.8–4.9*10⁹/L), increased lactate dehydrogenase (LDH, median: 386 IU/mL, range: 293–478 IU/mL), and elevated ferritin (median: 281 ng/mL, range: 145–488 ng/mL). IFN‑γ was measured in only 24 of the 48 patients, with a median concentration of 14.50 pg/mL (IQR: 6.9–22.9 pg/mL).

Of the 48 subjects, fever subsided after corticosteroid therapy in 24 patients and combination therapy with corticosteroid plus methotrexate (MTX) in 1 patient, following lymph node biopsy in 17 patients, and spontaneously without specific treatment in 6 patients; recurrent cervical lymphadenitis developed in 7 individuals. Further subgroup comparisons demonstrated that TB-Nil levels were significantly higher in patients receiving corticosteroid alone or corticosteroid plus MTX [median: 0.90 (IQR: 0.665–1.32)] than in the non-steroid group [median: 0.40 (IQR: 0.29–1.16), (P = 0.0418)]. No significant intergroup difference in TB‑Nil was observed between patients with recurrent disease and those achieving complete remission(P = 0.2608).

Comparison of nil values in patients with HNL, sJIA, KD and BL

To analyze the expression characteristics of IFN-γ levels in the nil tube of the TB-IGRA in patients with HNL, we measured TB-IGRA levels in all patients. No significant differences in TB-Nil levels were observed between definitive and clinical HNL groups(P = 0.5653).The median Nil value in HNL patients (0.82 IU/mL, IQR: 0.35–1.22) was significantly higher than in patients with sJIA (median 0.10 IU/mL, IQR: 0.03–0.24, p < 0.0001), KD (median 0.10 IU/mL, IQR: 0.02–0.18, p < 0.0001), and BL (median 0.09 IU/mL, IQR: 0.035–0.145, p < 0.0001) (Fig. 1).

Fig. 1.

The differences of TB-Nil levels between patients with HNL (N = 48), Systemic Juvenile Idiopathic Arthritis (sJIA, N = 31), Kawasaki Disease(KD, N = 21)and Bacterial Lymphadenitis (BL, N = 22). Data are expressed as medians, and the error bars indicate interquartile ranges. P < 0.05 represents the statistically significant

Diagnostic potential of the TB-Nil value

ROC analyses of TB-Nil were conducted for all HNL patients versus disease controls, as well as for the subgroup of pathologically confirmed definitive HNL (Fig. 2). The corresponding AUC values, 95% confidence intervals, sensitivities and specificities of these ROC analyses are summarized in Table 2.

Fig. 2.

Receiver operating characteristic (ROC) curves of TB-Nil for differentiating all HNL patients and pathologically confirmed definitive HNL from disease control subjects

Table 2.

ROC-derived AUC, 95%CI, sensitivity and specificity of TB-Nil for discrimination of total HNL and definitive HNL from control subjects

	AUC	Cut-off value	Sensitivity(%)	Specificity(%)	95%CI	P-value
ALL-HNL
All controls	0.88	0.19	91.67	78.38	0.81–0.94	<0.0001
sJIA	0.87	0.17	91.67	83.87	0.77–0.97	<0.0001
KD	0.87	0.19	91.67	68.18	0.78–0.96	<0.0001
BL	0.90	0.18	91,67	80.95	0.83–0.98	<0.0001
Definitive-HNL
All controls	0.86	0.19	89.19	78.38	0.79–0.94	<0.0001
sJIA	0.86	0.17	89.19	83.87	0.75–0.96	<0.0001
KD	0.85	0.19	89.19	68.18	0.75–0.95	<0.0001
BL	0.89	0.17	89.19	80.85	0.80–0.97	<0.0001

Comparison with plasma IFN-γ

Paired plasma samples for TB‑IGRA and IFN‑γ detection were collected concurrently in 24 of the 48 patients. Plasma IFN‑γ concentrations were quantified using a commercial human IFN‑γ enzyme‑linked immunosorbent assay (ELISA) kit in accordance with the manufacturer’s instructions. The median IFN‑γ level was 14.50 pg/mL (IQR: 1.15–315.9 pg/mL). Spearman’s rank correlation analysis was applied to assess the association between TB‑Nil and plasma IFN‑γ. No statistically significant correlation was identified (r = 0.1327, P = 0.5365).

Discussion

Tuberculosis infection should be excluded for patients with suspected histiocytic necrotizing lymphadenitis (HNL) [1]. Although the overall tuberculosis burden has gradually declined in China, tuberculosis remains a critical public health concern in certain endemic regions [17]. Similar to adults, children are susceptible to Mycobacterium tuberculosis infection, with children younger than 5 years at particularly high risk of developing severe, life-threatening tuberculous disease [18]. Atypical tuberculosis often presents with unexplained fever and lymphadenopathy without typical pulmonary manifestations, which frequently leads to missed or delayed diagnosis with fatal consequences [19]. Accordingly, TB‑IGRA testing is routinely performed to rule out tuberculosis in such patients. Previous pooled data report that QuantiFERON-TB Gold achieves a pooled sensitivity of 91%–94% and specificity of 97%–98% [12]. In the present cohort of 48 HNL patients, all subjects yielded negative TB‑IGRA results, accompanied by negative PPD skin tests, no documented tuberculous exposure history, and complete fever remission in the absence of anti‑tuberculosis treatment. By utilizing existing real-world clinical data, our findings further corroborate the high specificity of TB‑IGRA for excluding tuberculosis among HNL patients.

The pathogenesis of HNL is thought to be closely related to immune dysregulation, with abnormal activation of the interferon pathway playing a central role [20]. Previous studies have shown a prominent type I interferon signature in the lymph node tissue of HNL patients [21], with overproduction of IFN-α by plasmacytoid dendritic cells (pDCs) being a key pathogenic mechanism [22]. Furthermore, Some studies have confirmed that serum IFN-γ levels are significantly elevated in HNL patients [8], and an IFN-γ/IL-6 ratio > 0.45 aids in differentiating HNL from Kawasaki disease or systemic juvenile idiopathic arthritis [23]. Our study builds upon these findings, demonstrating that the Nil value from the TB-IGRA—representing the background IFN-γ release from peripheral blood lymphocytes without specific antigen stimulation-is sign

ificantly elevated in HNL patients. This finding aligns with previous studies and further supports the existence of persistent IFN-γ pathway activation in HNL patients.

However, no significant linear correlation was observed between TB-Nil and simultaneously measured plasma IFN‑γ in our HNL cohort in our study, indicating these two parameters carry distinct clinical implications. In fact, TB-Nil does not simply represent the direct measurement of circulating serum IFN‑γ. Fundamentally, TB-Nil quantifies baseline IFN‑γ spontaneously secreted by peripheral lymphocytes under antigen-free in vitro incubation at 37 ± 1 °C for 16–24 h [24], mirroring the spontaneous activation capacity of lymphocytes after ex vivo separation [25]. In contrast, plasma IFN‑γ mirrors the steady-state inflammatory cytokine concentration in vivo, modulated by complicated immune networks, cytokine catabolism and systemic clearance pathways [26]. Such inherent discrepancies confirm that TB-Nil possesses independent diagnostic value rather than merely serving as an auxiliary index for peripheral plasma IFN‑γ detection.

In summary, this study demonstrates that the Nil value from the TB-IGRA can serve as a practical, rapid, and non-invasive adjunctive indicator for the early differential diagnosis of pediatric HNL. The clinical value of this parameter is reflected in several aspects: First, the Nil value is derived from a routine TB-IGRA test, requiring no additional invasive procedures or special tests, and can be obtained immediately in clinical practice. Second, for cases with atypical clinical presentations, where lymph node biopsy is difficult to perform, or when rapid differentiation from infectious lymphadenitis, Kawasaki disease, or sJIA is needed, the Nil value can provide valuable diagnostic clues. Third, an elevated Nil value may reflect persistent activation of the IFN-γ pathway in HNL patients, providing a new clinical observational basis for understanding the immunopathological mechanisms of the disease.

Limitations

This study has several inherent limitations worthy of consideration. First, the overall sample size was limited, with smaller numbers in certain control subgroups such as BL, which may constrain statistical power for stratified analyses to some extent. Second, as a single-center retrospective investigation, inherent selection and recording biases could not be completely avoided. Notably, partial clinical and laboratory data including fever duration, ANC, ALC and ferritin were incompletely documented across control cohorts due to retrospective data restrictions, restricting our ability to supplement full baseline characteristics for all comparison groups and precluding multivariate logistic regression adjustment with comprehensive confounding factors. Third, serial longitudinal monitoring of TB-Nil throughout acute, remission and recurrent disease stages was not performed, and its association with disease activity remains to be clarified in future prospective cohorts. Subsequent multicenter prospective research with larger cohorts would help further verify the diagnostic utility of TB-Nil for HNL and explore its links with immune indicators including circulating cytokines and lymphocyte subsets, facilitating more solid clinical evidence for early diagnosis and individualized management of HNL.

Conclusion

The present study identified significantly elevated TB‑Nil values in pediatric HNL patients, which exhibited favorable diagnostic performance for distinguishing HNL from other pediatric inflammatory lymphadenopathy diseases. Moreover, the lack of correlation between TB‑Nil and circulating plasma IFN‑γ confirmed that TB‑Nil represents an independent ex vivo lymphocyte activation biomarker, rather than a simple reflection of systemic serum IFN-γ status. Prospective multi-center studies are warranted to validate these findings.

Author contributions

Q.W, C.X.F, and X.Y.W. conceived and designed the study. Q.W., Y.G., and K.L.L. collected and assembled the data. Q.W., X.Y.W., and Y.X. analyzed and interpreted the data. Q.W. drafted the manuscript. Y.X. and C.X.F. critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This retrospective study was approved by the Institutional Review Board of Guangzhou Women and Children’s Medical Center, Guangzhou Medical University. The study was conducted in accordance with the Declaration of Helsinki. Due to the retrospective nature of the study, the requirement for informed consent was waived by the IRB.

Consent for publication

Not applicable. this manuscript contains no individual person’s data in any form.

Competing interests

The authors declare no competing interests.