# Quantitative Bone SPECT/CT in Diabetic Foot Osteomyelitis: Diagnostic Performance Within-Patient Lesion–Contralateral Separation and Associations with Inflammatory Burden

## TL;DR

This study shows that quantitative SPECT/CT imaging can effectively detect bone infections in diabetic feet and correlates with inflammation levels.

## Contribution

The study introduces specific SUV thresholds for diagnosing diabetic foot osteomyelitis and links imaging data with inflammatory biomarkers.

## Key findings

- Quantitative SPECT/CT SUVmean showed excellent discrimination (AUC 0.961) for diabetic foot osteomyelitis.
- SUV values correlated with inflammatory biomarkers like CRP and ESR, with ESR × CRP showing the strongest association.
- MCH and aPTT showed inverse relationships with SUVs, suggesting broader systemic inflammation links.

## Abstract

Objective: We sought to assess the diagnostic performance of quantitative bone SPECT/CT standardized uptake values (SUVs) in diabetic foot osteomyelitis (DFO) and their associations with inflammatory biomarkers. Methods: We retrospectively reviewed 150 consecutive patients who underwent three-phase bone scintigraphy and foot SPECT/CT (November 2016–December 2024) for DFO before antibiotic treatment; 117 with complete imaging and laboratory data were analyzed. Lesion and contralateral SUVs (SUVmax, SUVmean) were compared. Receiver operating characteristic (ROC) curves were used to determine discrimination and optimal cut-offs (Youden index). Associations with biomarkers (C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), CRP/albumin (ESR × CRP) and hematologic/coagulation indices including mean corpuscular hemoglobin (MCH) and activated partial thromboplastin time (aPTT) were evaluated using Spearman correlations. Results: Lesion uptake and contralateral uptake were SUVmax 10.94 ± 7.36 vs. 3.62 ± 1.70; SUVmean 4.38 ± 3.63 vs. 0.93 ± 0.50. Discrimination was excellent; SUVmax was AUC 0.921 (cut-off 4.47; sensitivity 0.93; specificity 0.75) and SUVmean was AUC 0.961 (cut-off 1.49; sensitivity 0.91; specificity 0.89). CRP and ESR showed weak but consistent positive correlations with SUVs (ρ ≈ 0.25–0.30). The ESR × CRP value correlated most strongly (e.g., with SUVmean ρ = 0.35), and CRP/albumin showed a modest positive association. MCH (ρ ≈ −0.20) and aPTT (ρ ≈ −0.37) were inversely related. Conclusions: Quantitative SPECT/CT provides excellent lesion–contralateral discrimination in DFO. SUVs—particularly SUVmean—track inflammatory burden, supporting their use as practical quantitative adjuncts to clinical and laboratory assessment. Study-specific cut-offs are promising but require local validation.