Do not abandon the bone marrow biopsy yet in diffuse large B-cell lymphoma.

Abstract

TO THE EDITOR: We read with interest the updated guidelines that were recently published in Journal of Clinical Oncology on the role of imaging inthestagingandresponseassessmentof lymphoma. Thenew guidelines claim that focal [F]fluorodeoxyglucose (FDG) uptake within thebonemarrowishighlysensitiveforbonemarrowinvolvementinboth Hodgkin lymphoma and aggressive non-Hodgkin lymphoma, and may obviate the need for a bone marrow biopsy (BMB). Although we agree thatFDG–positronemissiontomography(PET)/computedtomography can replace BMB in Hodgkin lymphoma, we believe that this is not the case in non-Hodgkin lymphoma, including diffuse large B-cell lymphoma (DLBCL). Sevenoriginalstudiesthatwereincludedinarecentmeta-analysison thevalueofFDG-PET/CTfor thedetectionofbonemarrowinvolvement in DLBCL reported FDG-PET/CT to have a high patient-based sensitivity, ranging between 70% and 95%. However, all studies that were included in that meta-analysis used both BMB and follow-up FDGPET/CT studies as a reference standard. Although a positive BMB can be considered as definitive proof of bone marrow involvement, FDG decrease at follow-up may not be sufficient evidence to confirm pretreatment bone marrow involvement. Benign bone marrow lesions may also exhibit a decrease in FDG uptake after treatment, and one cannot confidently discriminate between the two entities without histologic proof. Thus, the actual diagnostic value of FDG-PET/CT in this setting is likely lower than previously quoted. Indeed, when BMB is used as the only reference standard, the sensitivity of FDG-PET/CT declines, as has been shown by recent studies that reported that FDG-PET/CT can be negative in up to 25% to 50% of patients with a positive BMB. Moving from a patient-based analysis to a local, head-to-head comparison between FDG-PET/CT and BMB in the posterior iliac crest, our preliminary data even show that FDG-PET/CT can be negative at the posterior iliac crest in up to 80% of patients with a positive BMB. Interestingly, this observation applies to both low and high tumor–volume bone marrow involvement, andtobothlarge-cell(concordant)andsmall-cell(discordant)bonemarrow involvement. Another important issue is that, although a positive BMB has been established as a major adverse prognostic marker in DLBCL, this is not the case for FDG-PET/CT–based bone marrow involvement. Of the four publishedstudiesontheprognostic implicationsofFDG-PET/CT–based bone marrow assessment, three studies showed that focal pathologic FDG uptake in the bone marrow did not have any prognostic value at all. These observations also support the notion that focally increased bone marrow FDG uptake is not pathognomonic for lymphomatous involvementinDLBCL.Notsurprisingly,therecentlydevelopedNational Comprehensive Cancer Network International Prognostic Index for DLBCL does not incorporate imaging-based bone marrow involvement and only accepts histologically confirmed bone marrow involvement for risk stratification. In conclusion, it is our opinion that FDG-PET/CT has not (yet) been proven to be able to replace BMB in DLBCL, both from a diagnostic and prognostic perspective. Until convincing scientific evidence emerges in favor of FDG-PET/CT, we believe that BMB should not be abandoned when evaluating the bone marrow in patients with newly diagnosed DLBCL.