Diagnosing Bone Metastases in Breast Cancer: A Systematic Review and Network Meta-Analysis on Diagnostic Test Accuracy Studies of 2-[F]FDG-PET/CT, F-NaF-PET/CT, MRI, Contrast-Enhanced CT, and Bone Scintigraphy.

This systematic review and network meta-analysis aimed to compare the diagnostic accuracy of 2-[F]FDG-PET/CT, F-NaF-PET/CT, MRI, contrast-enhanced CT, and bone scintigraphy for diagnosing bone metastases in patients with breast cancer. Following PRISMA-DTA guidelines, we reviewed studies assessing 2-[F]FDG-PET/CT, F-NaF-PET/CT, MRI, contrast-enhanced CT, and bone scintigraphy for diagnosing bone metastases in high-stage primary breast cancer (stage III or IV) or known primary breast cancer with suspicion of recurrence (staging or re-staging). A comprehensive search of MEDLINE/PubMed, Scopus, and Embase was conducted until February 2024. Inclusion criteria were original studies using these imaging methods, excluding those focused on AI/machine learning, primary breast cancer without metastases, mixed cancer types, preclinical studies, and lesion-based accuracy. Preference was given to studies using biopsy or follow-up as the reference standard. Risk of bias was assessed using QUADAS-2. Screening, bias assessment, and data extraction were independently performed by two researchers, with discrepancies resolved by a third. We applied bivariate random-effects models in meta-analysis and network meta-analyzed differences in sensitivity and specificity between the modalities. Forty studies were included, with 29 contributing to the meta-analyses. Of these, 13 studies investigated one single modality only. Both 2-[F]FDG-PET/CT (sensitivity: 0.94, 95% CI: 0.89-0.97; specificity: 0.98, 95% CI: 0.96-0.99), MRI (0.94, 0.82-0.98; 0.93, 0.87-0.96), and F-NaF-PET/CT (0.95, 0.85-0.98; 1, 0.93-1) outperformed the less sensitive modalities CE-CT (0.70, 0.62-0.77; 0.98, 0.97-0.99) and bone scintigraphy (0.83, 0.75-0.88; 0.96, 0.87-0.99). The network meta-analysis of multi-modality studies supports the comparable performance of 2-[F]FDG-PET/CT and MRI in diagnosing bone metastases (estimated differences in sensitivity and specificity, respectively: 0.01, -0.16 - 0.18; -0.02, -0.15 - 0.12). The results from bivariate random effects modelling and network meta-analysis were consistent for all modalities apart from F-NaF-PET/CT. We concluded that 2-[F]FDG-PET/CT and MRI have high and comparable accuracy for diagnosing bone metastases in breast cancer patients. Both outperformed CE-CT and bone scintigraphy regarding sensitivity. Future multimodality studies based on consented thresholds are warranted for further exploration, especially in terms of the potential role of F-NaF-PET/CT in bone metastasis diagnosis in breast cancer.