Abstract Infectious spondylodiscitis can cause progressive vertebral destruction and instability even after infection control, and some patients decline or are unsuitable for surgery. We report a 67-year-old man with poorly controlled type 2 diabetes mellitus and hypertension who developed Escherichia coli bacteremia and subsequent infectious spondylodiscitis. Four months after treatment of a traumatic open tibial wound with debridement and split-thickness skin grafting, he presented with progressive low back pain, intermittent fever, and systemic illness with acute kidney injury and hepatic dysfunction. A distal common bile duct stone was treated endoscopically, but fever persisted. A gallium-67 scan raised suspicion for L2–L3 involvement, and lumbar spine magnetic resonance imaging demonstrated L2–L3 infectious spondylodiscitis with extensive paraspinal and epidural inflammatory changes and bilateral psoas abscesses. Computed tomography-guided aspiration and bone biopsy were culture-negative. He improved clinically after approximately five weeks of antimicrobial therapy and declined operative debridement or stabilization. One week after completing antibiotics, denosumab (60 mg, subcutaneous) was administered. Given that denosumab is not a guideline-supported therapy for infectious spondylodiscitis, it was used off-label as a speculative adjunct to suppress inflammation-associated bone resorption; a second dose was given six months later. Back pain and function improved with bracing, and serial radiographs showed new bone formation along the superior endplate of L3 with progressive remodeling. At six-year follow-up, the patient remained free of adverse effects and demonstrated spontaneous fusion of L2 and L3 with sustained symptom relief. This report is descriptive and hypothesis-generating. The observed remodeling may also reflect expected post-infectious healing under conservative management. Nevertheless, the course raises a testable hypothesis that carefully timed antiresorptive therapy after confirmed infection control may support structural recovery in selected nonoperative patients with marked infection-related vertebral osteolysis.
Lu et al. (Thu,) studied this question.