Pretransplant FDG-PET in aggressive non-Hodgkin lymphoma: systematic review and meta-analysis.

This study aimed to systematically review and meta-analyze the value of pretransplant FDG-PET in predicting outcome after autologous stem cell transplantation in aggressive non-Hodgkin lymphoma. MEDLINE was systematically searched; included studies were methodologically assessed and meta-analyzed, when possible. Overall methodological quality of included studies (n = 11) was poor, with moderate risk of bias in the domains of study participation (n = 7) and prognostic factor measurement (n = 7), and high risk of bias in the domains of outcome measurement (n = 10), and study confounding (n = 11). In all aggressive non-Hodgkin lymphomas, pooled sensitivity and specificity were 54.0% and 73.1% in predicting treatment failure, and 54.5% and 68.7% in predicting death. Because of interstudy heterogeneity, additional subgroup analyses were performed. In newly diagnosed aggressive non-Hodgkin lymphoma, pooled sensitivity and specificity were 20.0% and 70.0% in predicting treatment failure, and 8.3% % and 30.5% in predicting death. In refractory/relapsed aggressive non-Hodgkin lymphoma, pooled sensitivity and specificity were 68.1% and 72.1% in predicting treatment failure, and 77.3% and 69.6% in predicting death. At present, pretransplant FDG-PET cannot be recommended in aggressive non-Hodgkin lymphoma, because available studies suffer from major methodological flaws, and reported prognostic estimates are low (i.e., poor in newly diagnosed and moderate in refractory/relapsed aggressive non-Hodgkin lymphoma).