Plasma microbial cell-free DNA sequencing for detection of Coxiella burnetii aortic endograft infection with vertebral osteomyelitis

Abstract

Chronic Q fever can present as culture-negative aortic endograft infection with osteomyelitis. Plasma microbial cell-free DNA sequencing unexpectedly detected Coxiella burnetii before serology, prompting exposure reassessment and targeted therapy. Later operative cultures grew Eikenella corrodens, underscoring the need to pair molecular results with serology and source-control microbiology.

Introduction

Coxiella burnetii is a zoonotic intracellular gram-negative bacterium transmitted by aerosols from infected livestock. ¹ Chronic infection may involve prosthetic valves, vascular grafts, aneurysms, or bone. ¹ Vascular infection with vertebral osteomyelitis has been reported. ^2,3 Diagnosis may be missed because symptoms are indolent, blood cultures are typically negative, and exposure history may not be elicited early. Serology remains the diagnostic standard, while PCR is specimen- and timing-dependent. ¹ Plasma microbial cell-free DNA sequencing (mcfDNA-seq) is a noninvasive adjunct for selected culture-negative infections when the differential includes fastidious organisms. ^4,5 Prior reports describe mcfDNA-seq detection of C. burnetii in culture-negative endocarditis and acute Q fever. ^6,7 We present a case in which mcfDNA-seq unexpectedly redirected evaluation toward chronic Q fever, while emphasizing confirmatory testing and limitations.

Clinical case

A 69-year-old man from Kansas with prior thoracoabdominal aortic aneurysm repair was transferred on hospital day 0 for suspected aortic endograft infection after 1 year of back pain and several months of chills. His history included rheumatoid arthritis treated with prednisone 5 mg daily and tobacco use. Before transfer, he had undergone outpatient spine injections and a nerve block. At an outside hospital on day-7, L1 bone biopsy showed fibrosis/chronic inflammation without malignancy; blood cultures and bacterial/acid-fast bacilli cultures from bone biopsy and psoas abscess pus had no growth. He took amoxicillin for a dental abscess before transfer. After the mcfDNA-seq result, a focused exposure history revealed work in an oil field, residence near cattle, and walks near cattle and sheep excretions.

Magnetic resonance imaging showed thoracolumbar discitis/osteomyelitis from T10 through L2 with T12 and L1 compression fractures and paraspinal/psoas inflammatory signal (Figure 1(A)). F-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) demonstrated intense linear uptake along the thoracoabdominal endograft, hypermetabolic paraspinal soft tissue involvement extending from T10 through L4, and intense uptake in the left psoas muscle corresponding to abscess formation (Figure 1(B)). These findings supported aortic endograft infection with contiguous osteomyelitis and psoas abscess.

Figure 1.

Figure 1 long description.

Multimodal imaging of aortic endograft infection with contiguous vertebral osteomyelitis and psoas involvement. (A) Sagittal short tau inversion recovery magnetic resonance image of the thoracolumbar spine shows abnormal marrow signal from T10 through L2 with endplate irregularity, disc involvement, T12 and L1 compression fractures, and paraspinal/psoas inflammatory signal. (B) Coronal fused F-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) image shows linear hypermetabolic activity along the thoracoabdominal aortic endograft, paraspinal soft tissue involvement, vertebral body uptake, and focal left psoas uptake. (C) Sagittal fused FDG PET/CT image shows hypermetabolic activity along the aortic endograft and adjacent anterior vertebral bodies. (D) Adjacent sagittal fused FDG PET/CT image shows hypermetabolic thoracolumbar vertebral and paraspinal soft tissue uptake. (E) More lateral sagittal fused FDG PET/CT image shows hypermetabolic uptake extending to the psoas compartment. Physiologic uptake is present in the brain, renal collecting systems, and urinary bladder.

Vancomycin and piperacillin-tazobactam were started empirically. Q fever was not the working diagnosis when plasma mcfDNA-seq (Karius Test; Karius, Inc., Redwood City, CA) was sent on day 1 for culture-negative endograft infection. On day 3, mcfDNA-seq detected low-level C. burnetii (<50 molecules/microliter), prompting doxycycline plus hydroxychloroquine, exposure reassessment, and serology (Supplementary Table 1). Q fever serology sent on day 5 resulted on day 8: phase I IgG 1:65,536, phase II IgG 1:4,096, and phase I/II IgM negative. With compatible imaging, this met Dutch consensus criteria for proven chronic Q fever vascular infection. ⁸ C. burnetii PCR was not performed on blood or tissue. Transesophageal echocardiography showed no vegetation or evidence of endocarditis.

On day 27, he underwent explantation of infected thoracoabdominal endografts, debridement, and graft reconstruction. Operative findings included infected aortic graft material, psoas abscess, T10-L2 osteomyelitis, and empyema. Operative cultures grew Eikenella corrodens; ceftriaxone was given for 10 days while doxycycline-hydroxychloroquine continued. He later developed acute-on-chronic subdural hematoma and hypotension requiring vasopressors and died on day 48.

Diagnostic stewardship implications

This case is not an example of inaccessible specimens or a stand-alone mcfDNA-seq diagnosis. Its incremental role was diagnostic reframing: C. burnetii was not considered during the prior evaluation, and the mcfDNA-seq result preceded serology by 5 days, prompting exposure reassessment and pathogen-directed therapy. The result did not replace serology, targeted PCR or sequencing, echocardiography, operative cultures, or surgical source control. The later E. corrodens result from local source-control specimens should be interpreted as complementary, temporally distinct microbiology rather than a direct measure of day-1 plasma-test sensitivity.

Stewardship should begin with the syndrome and intended actionability. For suspected chronic Q fever, targeted serology remains first-line and provides titers for follow-up; blood or tissue PCR can confirm infection when positive. Plasma mcfDNA-seq may be defensible when cultures are negative, prior antibiotics may reduce yield, the differential remains broad, and a fastidious diagnosis would change evaluation or treatment. However, it is costly relative to targeted testing, Q fever-specific clinical performance data are limited, and false-positive or clinically irrelevant detections can mislead care. ⁹ Results require correlation with serology, imaging, cultures, operative findings, and infectious diseases consultation (Table 1).

Table 1.

Diagnostic stewardship comparison for suspected chronic Q fever with vascular graft or vertebral involvement

Diagnostic modality	Best use	Turnaround/performance/cost considerations	Limitations and stewardship guardrails
Serology (phase I/II IgG and IgM)	First-line confirmatory test for suspected chronic Q fever; phase I IgG titers support classification and follow-up.	Turnaround is laboratory-dependent, often several days; generally less costly than broad plasma mcfDNA-seq.	Interpret with a compatible clinical focus; IgM may be negative in chronic disease; no organism recovery or co-pathogen detection.
PCR from blood, serum, or tissue	Confirmatory when positive, especially from operative tissue or other infected sites; blood/serum can be attempted.	Specificity is high, but blood/serum sensitivity is limited in chronic infection; turnaround depends on reference laboratory.	Specimen- and timing-dependent; negative result does not exclude Q fever; C. burnetii PCR was not performed in this case.
Plasma mcfDNA-seq (eg, Karius)	Adjunct for broad, culture-negative deep infection when a fastidious diagnosis would change evaluation or therapy.	May return faster than some send-out tests after specimen receipt; higher cost than targeted testing; Q fever-specific clinical performance is uncertain.	False-positive or clinically irrelevant detections; no susceptibility data; does not replace operative cultures, which may identify organisms later from local source-control specimens.

C. burnetii, Coxiella burnetii; IgG, immunoglobulin G; IgM, immunoglobulin M; mcfDNA-seq, microbial cell-free DNA sequencing; PCR, polymerase chain reaction.

Discussion

Chronic Q fever should be considered in culture-negative vascular graft infection or aneurysm infection, particularly with back pain, psoas involvement, immunosuppression, and livestock exposure. FDG PET/CT may localize C. burnetii vascular graft infection, but imaging establishes anatomy rather than etiology. ¹⁰ In this case, serology and imaging confirmed the mcfDNA-seq signal; absence of C. burnetii PCR from blood or tissue limits assessment of organism burden at the operative site. The fatal outcome precludes outcome-benefit claims.

Advanced diagnostics are most useful when ordered with a specific management question and interpreted conservatively. In suspected chronic Q fever with endovascular or vertebral involvement, plasma mcfDNA-seq may accelerate recognition when Q fever has not yet been considered, but it should be paired with confirmatory serology and tissue-based testing whenever possible. This approach preserves the diagnostic benefit of molecular testing while minimizing unnecessary antimicrobial exposure, false-positive or irrelevant detections, and low-yield testing.

Supporting information

Jain et al. supplementary material

Jain et al. supplementary material

Figure 1. Long description

Panel A: Sagittal short tau inversion recovery magnetic resonance image of the thoracolumbar spine shows abnormal marrow signal from T10 through L2 with endplate irregularity, disc involvement, T12 and L1 compression fractures, and paraspinal/psoas inflammatory signal. Panel B: Coronal fused F-18 fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) image shows linear hypermetabolic activity along the thoracoabdominal aortic endograft, paraspinal soft tissue involvement, vertebral body uptake, and focal left psoas uptake. Panel C: Sagittal fused FDG PET/CT image shows hypermetabolic activity along the aortic endograft and adjacent anterior vertebral bodies. Panel D: Adjacent sagittal fused FDG PET/CT image shows hypermetabolic thoracolumbar vertebral and paraspinal soft tissue uptake. Panel E: More lateral sagittal fused FDG PET/CT image shows hypermetabolic uptake extending to the psoas compartment. Physiologic uptake is present in the brain, renal collecting systems, and urinary bladder.

Navigate back to Figure 1..

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/ash.2026.10782.

Data availability statement

Deidentified clinical details supporting this case report are not publicly available because of patient privacy considerations.

Author contribution

Conceptualization: A.J., M.A.A. Data curation: A.J. Formal analysis: A.J. Investigation: A.J., S.K., R.S., A.H., A.A.F. Methodology: T.L. Project administration: M.A.A. Supervision: M.A.A. Visualization: A.A.F. Writingoriginal draft: A.J. Writing—review and editing: T.L., S.K., R.S., A.H., A.A.F., M.A.A. All authors reviewed and approved the final manuscript.

Financial support

None reported.

Competing interests

S.K. and M.R.S. report honoraria or consulting fees from Karius Inc. A.J., A.H., T.L., A.A.F., and M.A.A. report no conflicts of interest relevant to this article.

Patient consent

This case report was reviewed by the Baylor College of Medicine Institutional Review Board and determined to meet criteria for exemption; separate patient consent for publication was not required. No identifiable patient information is included.