Bone Marrow Biopsy Is Still Indispensable for Evaluating Bone Marrow Involvement in DLBCL Patients Even in the PET/CT Era

Background: Initial staging by positron emission tomography/computed tomography (PET/CT) is recommended for lymphoma of fluorodeoxyglucose (FDG)-avid histology, such as diffuse large B-cell lymphoma (DLBCL). Although both PET/CT and bone marrow biopsy (BMB) are usually performed to determine the bone marrow involvement in DLBCL, whether PET/CT alone is sufficient remains controversial. While previous reports discussing this point have used data based on medical records, here we performed a systematic review of untreated DLBCL patients, for whom a radiologist and a hematopathologist separately performed blind reviews. Methods: The subjects had DLBCL diagnosed between 2008 and 2017 for which both PET/CT and BMB were performed for staging before initiation of treatment. We excluded cases in which BMB specimens were qualitatively and/or quantitatively insufficient to determine the presence or absence of bone marrow involvement. To exclude follicular lymphoma, we did not include patients whose diagnostic specimens of lymphoma included any degree of nodular growth pattern. All PET/CT scans were performed in a single institution. In PET/CT, cases with score 4 or 5 in the Deauville criteria in bone marrow were classified as positive. In the review of BMB, we used both hematoxylin-eosin staining and CD20 immunostaining specimens in all cases to determine the involvement. We treated the presence of bone marrow involvement by BMB as a reference, and the involvement type was classified as either concordant or discordant pattern. Survival curves were compared by the log-rank test. A P-value less than 0.05 was considered to indicate a significant difference. Results: The study included 87 untreated patients with DLBCL (82 DLBCL, NOS; 3 intravascular large B-cell lymphoma; and 2 primary mediastinal large B-cell lymphoma), comprising 46 males and 41 females. The median age at the time of diagnosis was 71 (range, 19 to 87 years). The number of positive cases in PET/CT and BMB was 17 (20%) and 24 (28%), respectively (Table). Nine (10%) cases had both positive results. Among BMB-positive cases, there were18 concordant and 6 discordant involvements. When considering BMB results as a reference, PET/CT showed 38% sensitivity and 87% for specificity. The positive and negative predictive values were 53% and 79%, respectively. In 61 patients initially treated with R-CHOP, there was no significant difference in progression-free survival (PFS) between positive and negative cases in PET/CT (P = 0.16) , but PFS was significantly worse for BMB-positive cases than BMB-negative cases (P = 0.03). Conclusion: BMB is still mandatory in patients with untreated DLBCL and predicts outcome in the era of PET/CT. Routine CD20 immunostaining might be helpful in detecting lymphoma involvement in BMB specimens. Disclosures No relevant conflicts of interest to declare.