Diagnostic Accuracy of Dual Energy CT for Bone Marrow Edema of the Sacroiliac Joints Compared with MRI: A Systematic Review

Background: Dual-energy computed tomography (DECT) with virtual non-calcium (VNCa) reconstruction is emerging as an alternative method for identifying bone marrow edema (BME) in cases of inflammatory sacroiliitis, especially when magnetic resonance imaging (MRI) is contraindicated or unavailable. Objectives: The aim of this systematic review and meta-analysis is to evaluate the diagnostic performance of DECT with VNCa reconstruction for detecting BME in inflammatory sacroiliitis, using MRI as the reference standard. Methods: Following PRISMA-DTA guidelines and a pre-registered PROSPERO protocol (CRD420251103652), we searched PubMed, Web of Science, and the Cochrane Library up to June 2025. Seven studies comprising 358 patients and 591 sacroiliac joints were included. Quality assessment was performed using QUADAS-2. Pooled sensitivity and specificity were calculated using a bivariate random-effects regression model. Heterogeneity was evaluated using the I2 statistic. Subgroup analyses were performed for anatomical site and slice thickness, along with a meta-regression according to anatomical site. Publication bias was assessed using Deeks’ test. Results: Quality assessment revealed a moderate risk of bias, primarily related to patient selection. The pooled sensitivity and specificity of DECT VNCa were 78% (95% CI: 65–88%) and 83% (95% CI: 71–91%), respectively. The area under the SROC curve was 0.91 (95% CI: 0.88–0.93), indicating good diagnostic accuracy. Heterogeneity was moderate for sensitivity (I2 ≈ 68%) and high for specificity (I2 ≈ 88%), largely driven by one outlier study. Sensitivity was lower for sacral BME than for iliac BME (69% vs 79%), although this difference did not reach statistical significance (p ≈ 0.07). Specificity tended to be higher with slice thickness ≥1 mm than with <1 mm (90% vs 84%), although the difference was not statistically significant (p ≈ 0.06). Exclusion of one outlier study (Deppe et al.) increased pooled specificity to 92% and reduced heterogeneity. Deeks’ test did not reveal significant publication bias. Conclusions: DECT VNCa has been shown to be highly accurate in the diagnosis of inflammatory sacroiliac BME. Its high specificity makes it suitable for confirming inflammatory activity when an MRI scan is not feasible. However, its sensitivity is moderate and variable, with a trend towards lower values for sacral lesions. This precludes its use as a standalone test to rule out active inflammatory sacroiliitis, especially in young patients. Standardization of acquisition protocols and site-specific diagnostic thresholds is recommended for optimal clinical implementation. Clinical Impact: Clarifying the diagnostic performance of VNCa of DECT to define its role as a complementary tool to MRI for assessing BME in inflammatory sacroiliitis. These results could guide appropriate patient selection and highlight the need for standardized protocols.