ABSTRACT
PURPOSE: Whole-body magnetic resonance imaging (WB-MRI) is a radiation-free alternative to the 99mTc-HDP bone scan (BS) for the detection of bone metastasis. The major drawback is the long examination time and application of gadolinium enhancer. The aim of this study is to analyze (i) the performance of WB-MRI versus the BS and (ii) the diagnostic benefit of gadolinium (WB-MRI + Gd) compared to a non-enhanced protocol (NE WB-MRI). METHODS AND MATERIALS: 1256 eligible WB-MRI scans were analyzed retrospectively with a single inclusion criterion, a clinical 12-month follow-up or a biopsy as ground truth. N = 285 patients received both a WB-MRI and a BS within 12 months. All the patients were imaged with a coronal T1w and a STIR, and n = 528 (42%) received an additional T1w-mDixon with gadoteridol (0.1 mmol Gd-DTPA/kg). RESULTS: From 1256 eligible patients, n = 884 (70%) had breast cancer as a primary disease, n = 101(8%) prostate cancer, and n = 77(6%) lung cancer. The sensitivity (Se) and negative predictive value (NPV) of the WB-MRI was 98/99%, significantly higher compared to BS with 82/89%, P < 0.001 Mc Nemar's test. The specificity (Spe) and positive predictive value (PPV) of the WB-MRI and BS was 85/82% and 91/86%, respectively. The interobserver agreement between WB-MRI and BS was 71%, Cohen's kappa 0.42. Analysis of the added diagnostic value of gadolinium revealed Se/Spe/PPV/NPV of 98/93/92/98% for the NE WB-MRI and 99/93/85/100% for the WM-MRI + Gd, P > 0.05 binary logistic regression with Fischer's exact test. CONCLUSION: WB-MRI exceeds the sensitivity of BS without compromising the specificity, even after omitting the gadolinium enhancer.