FDG PET/CT may replace bone marrow biopsy for the evaluation of bone marrow involvement in selected mature T- and...

To systematically determine the role of FDG PET/CT for the diagnosis of bone marrow involvement in mature T- and natural killer (NK)-cell lymphomas.

The PubMed, Embase and Cochrane Library databases were searched to identify eligible studies. Data extraction and quality assessment were independently conducted. Then, pooled diagnostic performance with the 95 % confidence interval (CI) was calculated and further analyzed based on different interpretation criteria, tumor type and stage.

Fifteen studies were eventually included for quantitative analysis. Overall, the methodological quality of included studies was acceptable. For detecting bone marrow involvement, FDG PET/CT achieved a poor sensitivity of 0.62 (95 % CI, 0.48-0.71) and a reasonable specificity of 0.92 (95 % CI, 0.87-0.96). Similar performance was observed for the specific type of extranodal NK/T-cell lymphoma (ENKTCL). In early-stage patients revealed by PET/CT, extremely small proportion (2/777) showed positive bone marrow biopsy, especially for the specific type of ENKTCL, whereas in advanced-stage patients, the specificity of FDG PET/CT dropped to 0.77 (95 % CI, 0.72-0.82). Regarding the interpretation, both diffuse and focal increased uptake patterns as positivity may result in increased sensitivity but decreased specificity compared with focal pattern alone as positivity.

FDG PET/CT demonstrated excellent negative predictive value for detecting marrow involvement in early-stage patients with mature T- and NK-cell lymphomas, especially the ENKTCL. Conversely, FDG PET/CT showed poor performance for the diagnosis of bone marrow involvement in advanced-stage patients.