Objective: To perform a meta-analysis comparing the diagnostic value of (18)FDG-PET, MRI, and bone scintigraphy (BS) in detecting bone metastases in patients with breast cancer.
Materials and methods: MEDLINE, EMBASE, Scopus, ScienceDirect, SpringerLink, Web of Knowledge, EBSCO, and the Cochrane Database of Systematic Review databases were searched for relevant original articles published from January 1995 to January 2010. Inclusion criteria was as follows: (18)FDG-PET, MRI or (99m)Tc-MDP BS was performed to detect bone metastases (the number of published CT studies was inadequate for meta-analysis and therefore could not be included in this study); sufficient data were presented to construct a 2 × 2 contingency table; histopathological analysis and/or close clinical and imaging follow-up for at least 6 months were used as the reference standard. Two reviewers independently assessed potentially eligible studies and extracted relevant data. A software program called "META-DiSc" was used to obtain the pooled estimates for sensitivity, specificity, diagnostic odds ratio (DOR), summary receiver operating characteristic (SROC) curves, and the *Q index for each modality.
Results: Thirteen articles consisting of 23 studies fulfilled all inclusion criteria. On a per-patient basis, the pooled sensitivity estimates for MRI (97.1%) were significantly higher than those for PET (83.3%) and BS (87.0%; P <0.05). There was no significant difference between PET and BS (P <0.05). The pooled specificity estimates for PET (94.5%) and MRI (97.0%) were both significantly higher than those for BS (88.1%; P <0.05). There was no significant difference between PET and MRI (P >0.05). The pooled DOR estimates for MRI (298.5) were significantly higher than those for PET (82.1%) and BS (49.3%; P <0.05). There was no significant difference between PET and BS (P >0.05). The SROC curve for MRI showed better diagnostic accuracy than those for PET and BS. The SROC curve for PET was better than that for BS. The*Q index for MRI (0.935), PET (0.922), and BS (0.872) showed no significant difference (P ≥0.05). On a per-lesion basis, the pooled sensitivity estimates for BS (87.8%) were significantly higher than those for PET (52.7%; P <0.05). The pooled specificity estimates for PET (99.6%) were significantly higher than those for BS (96.1%; P <0.05).The pooled DOR estimates for PET (283.3) were significantly higher than those for BS (66.8%; P <0.05). The SROC curve for PET showed better diagnostic accuracy than that for BS. The*Q index for PET (0.941) was significantly higher than that for BS (0.893; P <0.05).
Conclusion: Magnetic resonance imaging was found to be better than (18)FDG-PET and BS for diagnosis of bone metastases in patients with breast cancer on a per-patient basis. On a per-lesion basis, (18)FDG-PET had lower sensitivity, higher specificity, a higher DOR, and a higher *Q index than BS.