Metagenomic Next-Generation Sequencing in Brucellar Spondylitis with Discordant Serologic Findings in a Non-Endemic Region: A Case Report and Literature Review

Abstract

Background

Brucellar spondylitis (BS) is a focal osteoarticular manifestation of brucellosis with nonspecific clinical and imaging features. Diagnosis may be difficult in areas where brucellosis is not traditionally considered endemic, particularly when epidemiologic exposure is not identified and serologic tests yield discordant results. We report a case in which metagenomic next-generation sequencing (mNGS) of vertebral lesion tissue provided additional microbiologic evidence for the diagnosis.

Case Presentation

A 71-year-old man presented with a 2-month history of fatigue, anorexia, night sweats, and worsening back pain without an identifiable epidemiologic exposure. Imaging showed destructive T8–T9 spondylitis with intervertebral disc involvement, paravertebral soft-tissue swelling, and small abscess formation. CT-guided biopsy demonstrated inflammatory changes, while routine bacterial culture, lesion-based MTBC DNA/RNA testing, and mycobacterial culture were negative. Brucella-specific immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies were positive, whereas the standard tube agglutination test (SAT) titer was <1:25 and the Rose Bengal test (RBT) was negative. mNGS of vertebral lesion tissue detected Brucella spp. DNA. Serial serologic testing subsequently showed SAT titers of 1:200 and 1:400, with later conversion of the RBT to positive, providing further support for brucellar spondylitis. The patient received combination anti-brucellar therapy and underwent posterior thoracic fixation, spinal canal decompression, lesion debridement, and bone graft fusion because of extensive vertebral destruction and concern for spinal instability. At approximately 10 weeks after treatment initiation, while anti-brucellar therapy was ongoing, back pain had markedly improved.

Literature Review

Published studies suggest that mNGS may improve pathogen detection in selected spinal infections when conventional microbiologic testing is unrevealing. Reports of focal brucellosis also show that lesion-derived specimens may provide microbiologic evidence that complements peripheral serologic findings.

Conclusion

In this case, mNGS did not identify a previously unsuspected pathogen but added microbiologic evidence from the affected vertebral lesion to an otherwise discordant serologic profile. Tissue-based mNGS may be useful in selected patients with suspected BS when conventional serologic findings are discordant and culture is negative.

Background

Brucellar spondylitis (BS) is a focal osteoarticular manifestation of human brucellosis and may cause vertebral destruction, paravertebral lesions, and spinal instability. Its clinical and imaging features are often nonspecific and may overlap with those of other spinal infections, particularly tuberculous spondylitis (TS), making etiologic diagnosis difficult.¹ Diagnosis may be further delayed in areas where brucellosis is uncommon or not traditionally considered endemic, especially when there is no clear epidemiologic exposure.

The diagnosis of human brucellosis is based on clinical, epidemiologic, and laboratory findings. Guidance developed jointly by the Food and Agriculture Organization of the United Nations (FAO), World Health Organization (WHO), and World Organisation for Animal Health (WOAH) recommends the combined use of bacteriologic and serologic tests. The Rose Bengal test (RBT) is commonly used for screening, whereas the standard tube agglutination test (SAT) and other serologic assays provide additional diagnostic information.² Results from different Brucella serologic tests may occasionally be inconsistent because of differences in antibody responses and assay methods.³ When culture is negative and histopathologic findings are nonspecific, such discordant results may make the diagnosis more difficult even when one serologic test suggests brucellosis.

Metagenomic next-generation sequencing (mNGS) allows culture-independent detection of microbial nucleic acids in clinical specimens.^4,5 It has increasingly been used in infectious spondylitis when conventional microbiologic tests fail to identify a pathogen.⁶ When vertebral lesion tissue is tested, mNGS may provide microbiologic evidence directly from the site of infection. We report a case of BS in a patient from an area where brucellosis is not traditionally considered endemic and who had no identifiable epidemiologic exposure. Brucella-specific immunoglobulin G (IgG) and immunoglobulin M (IgM) antibodies were positive, but SAT and RBT results were discordant. mNGS detected Brucella spp. DNA in vertebral lesion tissue, and subsequent increases in Brucella agglutination titers provided further support for the diagnosis.

Case Presentation

A 71-year-old man presented with a 2-month history of progressive fatigue, anorexia, night sweats, and worsening back pain. On March 1, 2024, he was evaluated at a local hospital. Laboratory testing showed anemia (hemoglobin, 107 g/L) and marked hyperferritinemia (serum ferritin, 4286.4 ng/mL). Chest computed tomography (CT) showed bronchitic changes and emphysema. The major diagnostic and therapeutic events are summarized in Figure 1, and the key laboratory, microbiological, and serologic findings are presented in Table 1.

Figure 1.

Timeline of the patient’s clinical course, diagnostic evaluation, serologic evolution, anti-brucellar treatment, surgical intervention, and follow-up.

Abbreviations: MRI, magnetic resonance imaging; PET/CT, positron emission tomography/computed tomography; TS, tuberculous spondylitis; IgG, immunoglobulin G; IgM, immunoglobulin M; SAT, standard tube agglutination test; RBT, Rose Bengal test; mNGS, metagenomic next-generation sequencing; DNA, deoxyribonucleic acid; CT, computed tomography; RIF, rifampicin; DOX, doxycycline; IV, intravenous; LVX, levofloxacin; TMP-SMX, trimethoprim-sulfamethoxazole.

Table 1.

Key Laboratory, Microbiological, and Serologic Findings During the Clinical Course

Date	Clinical Phase	Key Laboratory and Microbiological Findings	Brucella-Specific Microbiologic and Serologic Findings
Mar 1, 2024	Initial evaluation at a local hospital	Hemoglobin, 107 g/L; serum ferritin, 4286.4 ng/mL
Mar 29, 2024	Extended diagnostic evaluation	ANA negative; anti-Sm positive; quantitative anti-dsDNA mildly elevated; qualitative anti-dsDNA negative
Apr 16–17, 2024	Evaluation of infectious spondylitis	ESR, 32 mm/h; vertebral biopsy showed inflammatory fibrous and granulation tissue; routine bacterial culture showed no growth; acid-fast staining and MTBC DNA/RNA testing of vertebral lesion tissue were negative; mycobacterial culture subsequently remained negative after 42 days
Apr 18, 2024	Etiologic evaluation and initiation of anti-brucellar therapy	–	IgG positive; IgM positive; SAT <1:25; RBT negative
Apr 24, 2024	Early treatment period	–	Lesion-tissue mNGS detected Brucella spp. DNA; quantitative sequencing data unavailable
May 3, 2024	During anti-brucellar therapy	–	IgG positive; IgM negative; SAT 1:200; RBT negative
May 25, 2024	Postoperative anti-brucellar therapy	–	IgG positive; IgM negative; SAT 1:400; RBT positive

Notes: An em dash indicates that no corresponding result was available at that time point. Vertebral lesion tissue was obtained during CT-guided biopsy on April 17, 2024, and the mNGS result was reported on April 24, 2024. Mycobacterial culture of the lesion specimen remained negative after 42 days of incubation. Only the qualitative mNGS result was available in the archived medical record; quantitative sequencing parameters were unavailable.

Abbreviations: ANA, antinuclear antibody; anti-dsDNA, anti-double-stranded DNA antibody; anti-Sm, anti-Smith antibody; DNA, deoxyribonucleic acid; ESR, erythrocyte sedimentation rate; IgG, immunoglobulin G; IgM, immunoglobulin M; mNGS, metagenomic next-generation sequencing; MTBC, Mycobacterium tuberculosis complex; RBT, Rose Bengal test; RNA, ribonucleic acid; SAT, standard tube agglutination test.

On March 29, 2024, he underwent outpatient evaluation at our institution because of persistent constitutional symptoms and marked hyperferritinemia. Autoimmune testing showed negative antinuclear antibodies, positive anti-Smith antibodies, and mildly elevated anti-double-stranded DNA antibodies on quantitative testing, whereas qualitative anti-double-stranded DNA testing was negative. In the absence of compatible clinical manifestations, these findings were insufficient to support a systemic autoimmune disease. Hematologic evaluation was also performed, but no hematologic malignancy was identified in the subsequent investigations available for this report.

A detailed epidemiologic history revealed no recognized risk factors for brucellosis. The patient denied livestock contact, occupational exposure to animals or animal products, consumption of unpasteurized dairy products, and recent travel to pastoral areas or regions recognized as endemic for brucellosis. With progressive back pain and evidence of destructive thoracic vertebral disease, infectious spondylitis became the leading diagnostic consideration.

On April 16, the patient was admitted for further evaluation and treatment. Laboratory testing showed a mildly elevated erythrocyte sedimentation rate of 32 mm/h. Physical examination revealed mild tenderness over the T8–T9 spinous processes. Motor strength and muscle tone in both lower extremities were preserved, sensation was intact, and knee and ankle reflexes were present bilaterally. Babinski signs were absent, with no evident neurologic deficit.

Thoracic spine magnetic resonance imaging (MRI) performed on April 16 showed abnormal signal intensity in the T8–T9 vertebral bodies, with hypointensity on T1-weighted images and heterogeneous hyperintensity on fat-suppressed T2-weighted images. Intervertebral disc involvement, indistinct vertebral margins, paravertebral soft-tissue swelling with small abscess formation, and mild spinal canal narrowing were also present. Positron emission tomography/CT performed on the same day showed osteolytic destruction of the T8–T9 vertebrae with increased fluorodeoxyglucose uptake and surrounding soft-tissue involvement. These findings were consistent with infectious spondylitis but were not specific for a particular pathogen. Given the thoracic vertebral destruction and paravertebral involvement, tuberculous spondylitis (TS) remained an important early differential diagnosis (Figure 2).

Figure 2.

Preoperative sagittal fat-suppressed T2-weighted magnetic resonance imaging of the thoracic spine. Red arrows indicate vertebral destruction and intervertebral disc involvement at T8–T9. The yellow arrow indicates the paravertebral abscess.

On April 17, CT-guided biopsy of the T8–T9 lesion was performed. Vertebral lesion tissue and purulent and hemorrhagic aspirate were obtained for histopathologic and microbiologic examination. Histopathology showed degenerated fibrocartilage and bone fragments with proliferation of inflammatory fibrous tissue and granulation tissue, without evidence of malignancy. Acid-fast, periodic acid–Schiff, periodic acid–methenamine silver, and Alcian blue stains were negative, and fluorescence staining for acid-fast bacilli was also negative. Combined DNA and RNA testing of the vertebral lesion tissue for the Mycobacterium tuberculosis complex (MTBC) was negative. Routine bacterial culture showed no growth, and mycobacterial culture of the lesion specimen subsequently remained negative after 42 days of incubation. Thus, conventional evaluation of the biopsy demonstrated an inflammatory process but did not establish the etiology.

On April 18, blood was collected for Brucella serologic testing. Brucella-specific IgG and IgM antibodies were positive, whereas the standard tube agglutination test (SAT) titer was <1:25 and the Rose Bengal test (RBT) was negative, resulting in discordant serologic findings. Anti-brucellar therapy was initiated on April 18 while etiologic evaluation was ongoing. The initial regimen consisted of rifampicin 450 mg once daily, doxycycline 100 mg twice daily, and intravenous levofloxacin 400 mg once daily.

mNGS of vertebral lesion tissue obtained during the April 17 biopsy subsequently detected Brucella spp. DNA, providing molecular evidence from the involved spinal lesion. Only the qualitative mNGS result was available in the archived medical record; detailed quantitative sequencing data could not be retrieved.

Serial serologic testing subsequently became more supportive of brucellosis. On May 3, the SAT titer had increased to 1:200. The RBT remained negative, Brucella-specific IgG remained positive, and IgM had become negative.

Given the extensive T8–T9 vertebral destruction, paravertebral abscess formation, and concern for spinal instability, surgical treatment was undertaken. On May 9, the patient underwent posterior thoracic pedicle screw fixation, spinal canal decompression, lesion debridement through a costotransverse approach, and bone graft fusion. Histopathologic examination of the surgical T8–T9 specimen showed fractured bony trabeculae with proliferation of inflammatory fibrous tissue and focal chronic suppurative inflammation. The same special stains were again negative, as was repeat fluorescence staining for acid-fast bacilli.

Levofloxacin was administered from April 18 to May 10 and was then discontinued, while rifampicin and doxycycline were continued. No transient worsening of clinical symptoms temporally associated with the initiation of levofloxacin was documented in the available medical records.

On May 25, repeat serologic testing showed a further increase in the SAT titer to 1:400. The RBT had become positive, Brucella-specific IgG remained positive, and IgM remained negative. The evolving serologic findings, together with detection of Brucella spp. DNA in vertebral lesion tissue and the compatible clinical and imaging features, supported the diagnosis of brucellar spondylitis. Meanwhile, the available lesion-based MTBC DNA/RNA testing, mycobacterial culture, acid-fast staining, and histopathologic examinations provided no evidence supporting mycobacterial infection.

The postoperative course was clinically stable. Incisional pain gradually improved, and no fever, wound drainage, or new lower-extremity neurologic deficit was documented before discharge. On May 27, the patient was discharged on rifampicin 450 mg once daily, doxycycline 100 mg twice daily, and trimethoprim-sulfamethoxazole 160/800 mg every 12 hours. Monthly laboratory monitoring and CT and MRI every 3 months were recommended, together with regular orthopedic follow-up.

At the last available clinical evaluation in late June 2024, approximately 10 weeks after initiation of anti-brucellar therapy, back pain had markedly improved, although mild activity limitation and occasional night sweats persisted. Follow-up CT showed stable postoperative changes at T8–T9. Anti-brucellar therapy was still ongoing at this last available follow-up.

No subsequent clinical records were available after June 2024. Therefore, the final duration of antimicrobial therapy, criteria for treatment discontinuation, subsequent serologic evolution, and late relapse status could not be determined.

Discussion

The reported geographic distribution of human brucellosis has expanded over the past two decades. The number of countries reporting human brucellosis has increased from approximately 53 to at least 97.⁷ In China, reported cases increased from 38,151 (2.8 per 100,000 population) in 2011 to 70,439 (5.2 per 100,000 population) in 2023, while the number of counties reporting cases increased from 834 to 2290.⁸ Although 96.1% of cases during 2011–2023 were reported in northern provincial-level administrative divisions, southern regions showed increasing incidence and geographic spread, with emerging clusters identified in Yunnan, Guangdong, and Xizang.⁸ Long-term spatiotemporal analyses have similarly shown expansion beyond traditionally high-incidence northern regions. Socioeconomic conditions, transportation, healthcare resources, and environmental factors have been associated with these changes, although such associations do not establish causality.⁹ These epidemiologic changes are clinically important because brucellosis may now be encountered in areas where clinicians are less likely to suspect it. Lack of an identifiable exposure history should therefore not by itself exclude brucellosis, particularly in patients with focal disease such as BS.

BS usually has a subacute or chronic course and commonly presents with nonspecific symptoms such as fever, sweating, fatigue, and back pain.¹⁰ Its clinical and radiologic features overlap with those of TS and other infectious spondylitides. In a retrospective study of 131 patients with infectious spondylodiscitis, including 63 with BS and 15 with TS, thoracic involvement was more common in TS than in BS (33.3% vs 11.1%), as was psoas abscess formation (46.7% vs 4.8%). Lumbar involvement was more common in BS (61.9% vs 26.7%).¹¹ Sweating was also more common in BS than in pyogenic spondylodiscitis.¹¹ These differences may help guide the differential diagnosis, but the considerable overlap between groups limits the usefulness of any single feature.

Several studies have attempted to improve the distinction between BS and TS by combining clinical, laboratory, and imaging findings. Abudukadier et al developed a nomogram based on symptom duration, anorexia, adenosine deaminase level, and psoas abscess formation in 183 patients with BS or TS. The model achieved areas under the receiver operating characteristic curve (AUCs) of 0.820 in the training cohort and 0.899 in the validation cohort.¹² CT findings may also be useful. Xu et al reported that fan-shaped osteosclerosis favored BS, with a positive predictive value (PPV) of 86.96%, whereas fragmented and blocked sequestra favored TS, with PPVs of 98.31% and 100%, respectively.¹³ These values apply to individual imaging signs and should not be interpreted as the overall diagnostic accuracy of CT.

Models that combine several types of information may perform better than individual findings. Zhu et al developed a multiclass random forest model using MRI and laboratory data to distinguish pyogenic, tuberculous, brucellar, and fungal spondylitis. The macro-average AUCs were 0.94 in the training set and 0.92 in the external test set, although sensitivities were only 0.61 and 0.66.¹⁴ Ma et al reported five noninvasive predictors for differentiating spinal tuberculosis from BS. Individual predictors had AUCs of 0.597–0.694, whereas the combined model achieved an AUC of 0.854; external validation was not reported.¹⁵ These studies show that clinical and imaging findings can improve diagnostic discrimination, but they cannot directly identify the causative organism.

Culture provides direct microbiologic evidence of brucellosis, but its yield is variable. Han et al reported blood culture positivity in 30.9% of 1188 hospitalized male patients with newly diagnosed, laboratory-confirmed brucellosis.¹⁶ In another study of 130 patients with acute or subacute brucellosis, 46.2% were blood culture positive and 53.8% were nonbacteremic.¹⁷ These findings show that bacteremia is not detected in all patients with brucellosis. However, because both studies evaluated blood culture, their results cannot be directly applied to vertebral tissue culture in BS. Culture yield may also be affected by disease stage, prior antimicrobial therapy, specimen type, and sampling conditions.^16,17 In our patient, routine bacterial culture of vertebral lesion tissue was negative.

Serologic testing remains a central part of the diagnosis of human brucellosis.² In a systematic review and meta-analysis of 38 studies, Freire et al reported a pooled sensitivity of 96.6% and specificity of 97.9% for RBT when culture and/or SAT were used as reference standards. Pooled sensitivity and specificity for IgG/IgM enzyme-linked immunosorbent assay were 96.8% and 98.6%, respectively.¹⁸ However, there was substantial variation among studies, and studies focused on complicated brucellosis were excluded. These estimates therefore may not apply directly to focal disease such as BS.¹⁸

Differences between serologic tests may also reflect the timing and type of antibody response. Yagupsky et al noted that SAT titers may be negative or below commonly used diagnostic thresholds early in the disease course and that antibody responses vary between patients.³ Low or false-negative SAT results may also occur because of the prozone phenomenon or predominance of nonagglutinating antibodies.³ Repeat testing may therefore be useful when brucellosis remains clinically suspected. Seroconversion or a fourfold or greater rise in antibody titer between paired specimens provides additional diagnostic support.³

Data specifically from patients with BS also show that SAT and RBT are not positive in every case. In a retrospective study of 200 patients with BS, SAT was positive in 157 of 172 tested patients (91.3%), and RBT was positive in 165 of 172 (95.9%).¹⁹ These data should be interpreted cautiously because of the retrospective design, the diagnostic criteria used, and the requirement for epidemiologic linkage. Nevertheless, they show that negative conventional serologic results can occur in patients diagnosed with BS.

In the present case, positive Brucella-specific IgG and IgM antibodies suggested brucellosis, but the initial SAT titer was <1:25 and the RBT result was negative. The patient also had no identifiable epidemiologic exposure, vertebral lesion culture was negative, and spinal imaging was not pathogen specific. The main diagnostic problem was therefore not a complete absence of evidence for brucellosis, but disagreement among the available findings. The later increase in Brucella agglutination titers from <1:25 to 1:200 and then to 1:400 provided additional serologic support.

TS was also considered because of the destructive T8–T9 lesions and paravertebral involvement. The available tuberculosis-related investigations and tissue findings did not provide microbiologic or histopathologic evidence of mycobacterial infection, although they did not completely exclude TS. At the same time, Brucella-specific IgG and IgM antibodies had already raised suspicion of brucellosis. Detection of Brucella spp. DNA in vertebral lesion tissue by mNGS therefore added microbiologic evidence directly from the affected site. The subsequent increase in agglutination titers further supported the diagnosis. In this case, the main contribution of mNGS was not to identify an entirely unexpected pathogen, but to provide a second type of etiologic evidence when the conventional serologic results were discordant.

Marked hyperferritinemia and abnormal autoantibody results complicated the initial diagnostic evaluation. Hyperferritinemia is nonspecific and may occur in infection, inflammation, liver injury, malignancy, immune-mediated disease, and disorders of iron metabolism.²⁰ The ferritin level of 4286.4 ng/mL therefore warranted further investigation but could not be attributed specifically to brucellosis. The abnormal autoantibody findings and other nonspecific hematologic abnormalities broadened the differential diagnosis; however, the available evaluation did not establish a systemic autoimmune disease or hematologic malignancy. These abnormalities cannot be directly linked to Brucella infection on the basis of this single case.

Evidence from other infectious diseases suggests that mNGS can provide useful additional diagnostic information. In a prospective multicenter study of 204 patients with meningitis or encephalitis, Wilson et al identified 58 nervous system infections. Thirteen infections (22%) were detected by cerebrospinal fluid mNGS but not by conventional clinical testing, whereas 26 were identified by conventional methods but not by mNGS.²¹ These findings support a complementary role for mNGS rather than replacement of conventional methods.

The principal studies evaluating mNGS in spinal infections and focal brucellosis are summarized in Table 2. Studies of spinal infection provide more directly relevant data. Shi et al evaluated 193 patients with suspected spinal infection, 162 of whom met the study reference standard for infection. mNGS had a sensitivity of 77.78%, compared with 27.16% for conventional microbial culture, with specificities of 90.32% and 96.77%, respectively.⁶ mNGS identified pathogens in 86 culture-negative cases, including Brucella and other uncommon organisms. Prior antimicrobial exposure was not significantly associated with mNGS detection, and positivity did not differ significantly between CT-guided biopsy and surgical specimens (80.00% vs 77.17%).⁶ These findings are relevant to our patient because mNGS was performed on CT-guided vertebral lesion tissue after routine culture was negative. However, the reported diagnostic performance depends on the reference standard used in that study.

Table 2.

Summary of Published Studies on mNGS in Spinal Infections and Focal Brucellosis

Sample Size	Infection Site	Country (Year)	Atypical Features	mNGS Result	Conventional Method Result	Ref.
193 (162 infected)	Spine	China (2025)	Culture-negative cases included	Sensitivity 77.78%; 86 culture-negative cases detected	Culture sensitivity 27.16%	[6]
27	Suspected spinal infection	China (2024)	Mixed spinal infections; 3 BS cases	Positive rate 85.19%	Conventional detection 48.15%	[22]
108 overall; mNGS, n=49; culture, n=39	Spine	China (2024)	Mixed spinal infections	Positive rate 61.2% (30/49); sensitivity 79.41%	Culture 30.8% (12/39); sensitivity 25.00%	[23]
1	Thoracic BS	China (2023)	Non-endemic region	Brucella spp. detected; 1 read matched B. melitensis	Blood/tissue cultures negative; RBT positive	[24]
1	Endophthalmitis	China (2022)	Rare focal infection	B. melitensis detected	Routine tests negative; SAT 1:25	[25]
1	Submandibular abscess	China (2025)	Rare-site infection	B. suis detected	SAT 1:200; blood culture positive	[26]
1	Central nervous system	China (2021)	Nonendemic setting	CSF mNGS positive; peripheral blood mNGS negative	CSF culture negative; SAT/qPCR supportive	[27]
29 studies	Spine	Systematic review (2026)	Mixed spinal infections	Detection 36.8–95.5%; sensitivity 39–94.7%	Culture detection 5.9–59.2%	[28]

Notes: Study designs, specimens, reference standards, and outcome measures differed across studies; reported detection rates and diagnostic performance should therefore not be directly compared.

Abbreviations: BS, brucellar spondylitis; CNS, central nervous system; CSF, cerebrospinal fluid; mNGS, metagenomic next-generation sequencing; qPCR, quantitative polymerase chain reaction; RBT, Rose Bengal test; SAT, standard tube agglutination test.

Li et al reported 27 patients with suspected spinal infection and found an mNGS positivity rate of 85.19%, compared with 48.15% for conventional pathogen detection; three patients had BS.²² Chen et al studied 108 patients with suspected spinal infection and also found a higher positive detection rate for mNGS than for conventional culture (61.2% vs 30.8%). mNGS had higher sensitivity than culture (79.41% vs 25.00%), whereas specificity and PPV did not differ significantly.²³ These studies suggest that mNGS may improve pathogen detection in some patients with spinal infection, particularly when conventional testing is unrevealing. However, positivity rates, detection rates, sensitivity, and specificity should not be treated as equivalent measures.

A recent systematic review included 29 studies of mNGS in spinal infection, including 25 retrospective studies and four case reports, all from China.²⁸ Reported pathogen detection rates ranged from 36.8% to 95.5% for mNGS and from 5.9% to 59.2% for culture, whereas mNGS sensitivity ranged from 39% to 94.7%.²⁸ In some studies, mNGS results were available within approximately 29–53 hours, compared with approximately 2–10 days for culture.²⁸ However, the included studies differed substantially in patient selection, specimen type, sequencing platform, reference standards, and definitions of positive results. The authors therefore did not provide a single pooled estimate of diagnostic performance. The fact that all studies were conducted in China also limits generalizability.

Several case reports have described the use of mNGS in focal Brucella infections. Du et al reported thoracic spinal infection caused by Brucella melitensis in a nonendemic region. Blood and tissue cultures were negative, whereas mNGS of CT-guided biopsy tissue detected Brucella-related sequences. A history of shepherding and a positive RBT result were identified after the mNGS result.²⁴ Only 20 reads were assigned to Brucella spp., with one read specifically matching B. melitensis, showing that low sequence counts may still be clinically relevant when supported by other findings.

Xi et al detected B. melitensis by mNGS of aqueous humor in a patient with infectious endophthalmitis after routine testing was negative. The serum SAT titer was 1:25, and frequent goat exposure was later identified.²⁵ Ao et al reported a submandibular abscess caused by Brucella suis, in which mNGS of pus was strongly positive and the diagnosis was also supported by livestock exposure, a SAT titer of 1:200, and positive blood culture.²⁶ Cao et al reported central nervous system brucellosis in which cerebrospinal fluid mNGS detected B. melitensis after an initially negative workup. The diagnosis was subsequently supported by quantitative polymerase chain reaction (qPCR), agglutination testing, and exposure history. Peripheral blood mNGS was negative while cerebrospinal fluid mNGS remained positive.²⁷

These reports cannot establish the diagnostic accuracy of mNGS, and the supporting evidence differed among cases. They do, however, show that Brucella can be detected by mNGS in specimens taken directly from affected sites, including vertebral tissue, aqueous humor, abscess material, and cerebrospinal fluid. The significance of an mNGS result therefore depends not only on detection of microbial sequences, but also on the specimen tested and whether the result is consistent with the clinical and laboratory findings.

mNGS also has technical and interpretive limitations. Background microbial sequences may arise from environmental or reagent contamination, normal microbial communities, or specimen processing.^4,5 Laboratories may also differ in sample preparation, sequencing platforms, reference databases, reporting thresholds, and bioinformatic methods.^4,5 Detection of microbial nucleic acid does not establish whether the organism is viable or whether it is responsible for the disease. mNGS results must therefore be interpreted together with the specimen source, clinical presentation, epidemiologic information, and conventional microbiologic findings. In this case, detailed quantitative mNGS data were unavailable. However, Brucella spp. DNA was detected directly in vertebral lesion tissue and was consistent with the positive Brucella-specific antibodies and the subsequent rise in agglutination titers.

Treatment of BS generally requires prolonged combination antimicrobial therapy, although the optimal regimen and duration have not been clearly established. Current literature describes two- or three-drug regimens that commonly include doxycycline, rifampicin, aminoglycosides, trimethoprim-sulfamethoxazole, or fluoroquinolones, with treatment usually lasting approximately 3–6 months and sometimes longer.¹ Surgery may be required in patients with neurologic compromise, epidural or large paravertebral abscesses, severe vertebral destruction or deformity, spinal instability, or poor response to antimicrobial treatment.¹ In our patient, extensive T8–T9 destruction, paravertebral involvement, and concern for instability led to surgical debridement, spinal canal decompression, posterior fixation, and fusion in addition to anti-brucellar therapy. Because long-term follow-up was incomplete, the final treatment duration and stopping criteria could not be assessed.

Limitations

This report has several limitations. First, vertebral lesion culture was negative; therefore, culture-based confirmation and species-level identification of Brucella were unavailable. Second, only the qualitative mNGS result was retained in the archived medical record. The sequencing platform, laboratory workflow, reporting thresholds, Brucella-specific read counts, relative abundance, genome coverage, and other quantitative parameters could not be retrieved. Third, DNA-based mNGS cannot distinguish viable from nonviable organisms and therefore cannot independently establish active infection. Fourth, this is a single-patient case report and cannot establish the diagnostic accuracy or incremental performance of mNGS relative to conventional testing. Finally, no clinical records were available after June 2024; therefore, the final treatment duration, criteria for treatment discontinuation, 6–12-month serologic evolution, and late relapse status could not be assessed.

Overall, this case shows that the diagnosis of BS can be difficult not only because its clinical and imaging features overlap with those of other spinal infections, but also because different diagnostic tests may provide discordant results. In this patient, Brucella-specific IgG and IgM antibodies provided an initial clue, whereas the SAT titer of <1:25 and negative RBT result were inconsistent with that finding. Detection of Brucella spp. DNA in vertebral lesion tissue by mNGS provided additional microbiologic evidence from the affected site, and the later rise in agglutination titers further supported the diagnosis. In patients with suspected BS and discordant Brucella serologic findings, tissue-based mNGS may be useful when conventional culture does not identify the pathogen.

Conclusion

Brucellar spondylitis should be considered in the differential diagnosis of destructive spinal infection, including in areas where brucellosis is not traditionally considered endemic and in patients without a recognized epidemiologic exposure. In this case, Brucella-specific IgG and IgM antibodies initially suggested brucellosis, but SAT and RBT results were discordant. Detection of Brucella spp. DNA by mNGS in vertebral lesion tissue provided additional microbiologic evidence from the affected site, and subsequent increases in agglutination titers further supported the diagnosis. Tissue-based mNGS may therefore be useful in selected patients with suspected BS when conventional serologic findings are discordant and culture is negative. Surgical treatment should be considered according to the extent of structural damage, abscess formation, neurologic involvement, spinal stability, and response to antimicrobial therapy.

Funding Statement

This study received no external funding.

Abbreviations

AUC, area under the receiver operating characteristic curve; BS, brucellar spondylitis; CT, computed tomography; DNA, deoxyribonucleic acid; ESR, erythrocyte sedimentation rate; IgG, immunoglobulin G; IgM, immunoglobulin M; mNGS, metagenomic next-generation sequencing; MRI, magnetic resonance imaging; MTBC, Mycobacterium tuberculosis complex; PPV, positive predictive value; qPCR, quantitative polymerase chain reaction; RBT, Rose Bengal test; RNA, ribonucleic acid; SAT, standard tube agglutination test; TS, tuberculous spondylitis.

Data Sharing Statement

The datasets generated and/or analyzed during the current study are not publicly available due to patient privacy considerations but are available from the corresponding author on reasonable request.

Ethics Approval and Consent to Participate

Ethical review and approval were waived by the Ethics Committee of The Third Affiliated Hospital of Zhejiang Chinese Medical University in accordance with local institutional requirements for single case reports.

Consent for Publication

Written informed consent was obtained from the patient for publication of this case report and any accompanying images.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.