PURPOSE

Low-virulent microorganisms identified on pedicle screws by sonication fluid culture (SFC) are an important cause of implant loosening. While sonication of explanted material improves the detection rate, the risk of contamination exists and no standardized diagnostic criteria for chronic low-grade spinal implant-related infection (CLGSII) are stablished. Besides, the role of serum C-reactive protein (CRP) and procalcitonin (PCT) in CLGSII has not been adequately investigated.

METHODS

Blood samples were collected prior to implant removal. To increase sensitivity, the explanted screws were sonicated and processed separately. Patients exhibiting at least one positive SFC were classified in the infection group (loose criteria). To increase specificity, the strict criteria only considered multiple positive SFC (≥ 3 implants and/or ≥ 50% of explanted devices) as meaningful for CLGSII. Factors which might promote implant infection were also recorded.

RESULTS

Thirty-six patients and 200 screws were included. Among them, 18 (50%) patients had any positive SFCs (loose criteria), whereas 11 (31%) patients fulfilled the strict criteria for CLGSII. Higher serum protein level was the most accurate marker for the preoperative detection of CLGSSI, exhibiting an area under the curve of 0.702 (loose criteria) and 0.819 (strict criteria) for the diagnosis of CLGSII. CRP only exhibited a modest accuracy, whereas PCT was not a reliable biomarker. Patient history (spinal trauma, ICU hospitalization and/or previous wound-related complications) increased the likelihood of CLGSII.

CONCLUSION

Markers of systemic inflammation (serum protein level) and patient history should be employed to stratify preoperative risk of CLGSII and decide the best treatment strategy.