A diagnostic model for differentiating tuberculous spondylodiscitis from pyogenic spondylodiscitis based on...

Area of Science:

• Spinal infections
• Radiology
• Infectious diseases

Background:

• Differentiating tuberculous spondylodiscitis (TS) from pyogenic spondylodiscitis (PS) is clinically challenging.
• Accurate differential diagnosis is crucial for appropriate treatment and patient outcomes.

Purpose of the Study:

• To distinguish TS from PS using laboratory, MRI, and CT findings.
• To develop a novel diagnostic model for the differential diagnosis of TS and PS.

Main Methods:

• Retrospective analysis of MRI, CT, and laboratory data from 81 patients (46 PS, 35 TS).
• Development of predictive models using binary logistic regression.
• Validation of the model through internal (bootstrap resampling) and external datasets.

Main Results:

• Key differentiating features for TS included disc signal/height preservation, skip lesions (≥3 segments), paravertebral calcification, massive sequestra, subligamentous bone destruction, osteosclerotic margins, higher WBC, and positive T-SPOT.TB.
• A four-predictor model (WBC<7.265, skip lesion/≥3 segments, massive sequestra, subligamentous bone destruction) demonstrated high accuracy (93.8%), sensitivity (91.4%), and specificity (95.7%).
• The model's predictive ability was confirmed by an AUC of 0.981 and consistent validation results.

Conclusions:

• A diagnostic model integrating CT, MRI, and laboratory findings effectively distinguishes TS from PS.
• This model offers a valuable tool for clinicians in diagnosing spinal infections.