Diagnostic accuracy of SPECT, PET, and MRS for primary central nervous system lymphoma in HIV patients: A systematic review and meta-analysis

Supplemental Digital Content is available in the text
Background: We performed a systematic review and meta-analysis to assess the roles of SPECT, PET, and MRS in distinguishing primary central nervous system lymphoma (PCNSL) from other focal brain lesions (FBLs) in human immunodeficiency virus (HIV)-infected patients. Methods: PubMed, Scopus, and Medline were systematically searched for eligible studies from 1980 to 2016. Two authors extracted characteristics of patients and their lesions using predefined criteria. Results: Eighteen studies on SPECT containing 667 patients, 6 studies on PET containing 108 patients, and 3 studies on MRS containing 96 patients were included. SPECT had a pooled sensitivity of 0.92 (95% CI: 0.85–0.96) and specificity of 0.84 (95% CI: 0.74–0.90) in differentiating PCNSL from other FBLs. For the 6 studies that used only pathology and/or serology as the gold standard, the pooled sensitivity was 0.85 (95% CI: 0.72–0.97) and the pooled specificity was 0.73 (95% CI: 0.54–0.92). Conclusion: SPECT has good diagnostic accuracy for discriminating PCNSL from other FBL-causing disorders in HIV patients. However, the actual sensitivity and specificity of SPECT may be lower than expected if only pathology and/or serology was used as the gold standard. PET may be superior but has less supporting clinical data and is more expensive.