Impact of Bone Marrow Evaluation Using Bone Marrow Biopsy, Flow Cytometry, and PET/CT in B-Cell Non-Hodgkin Lymphomas

Introduction The initial evaluation and staging of non-Hodgkin lymphomas (NHL) are currently conducted according to the Lugano classification (2014). The incorporation of PET/CT and other techniques such as flow cytometry (FCM) and molecular studies into routine clinical practice has recently questioned the utility of bone marrow biopsy (BMB) in the initial work-up of NHL. Our study aimed to evaluate bone marrow (BM) involvement through BMB and FCM, and also by PET/CT in patients with B-cell NHL. We assessed the impact of each method on lymphoma staging and their ability to discriminate progression-free survival (PFS) and overall survival (OS). Material and Methods This observational retrospective study consecutively included all patients with B-cell NHL of any histological type who underwent BM evaluation between 2019 and 2023 at a single tertiary hospital. Inclusion criteria were: > 18 years, B-cell NHL and BM evaluation through aspiration/trephine biopsy at diagnosis at our center. Diagnoses were grouped into indolent NHL (follicular lymphoma (FL), marginal zone lymphoma (MZL), or other indolent lymphomas) or aggressive NHL (diffuse large B-cell lymphoma (DLBCL), mantle cell lymphoma (MCL), or other aggressive lymphomas). The Ann Arbor (AA) stage was initially defined based on clinical and imaging findings by PET/CT without considering BM (AA-baseline), and subsequently incorporating BMB (AA-BMB), FCM (AA-FCM), and PET/CT considering BM findings (AA-PET). FCM was performed in BM aspirate according to EuroFlow based methods (at least eight colors) resulting in a general sensitivity between 10-3 and 10-4. Concordance Index (C-Index), Akaike Information Criterion (AIC), and estimated concordance probability (CPE) were used to evaluate discrimination for PFS and OS. Results A total of 224 patients were identified, of whom 134 (59%) met the inclusion criteria. The median age was 64 years (range 19-91) with 65.1% being male. Of the total number of patients, 219 (97.8%) were caucasian, 4 (1.8%) were arab, and 1 (0.4%) was black. Histologies were as follows: FL 25.4%, MZL 19.4%, other indolent NHL 2.2%, DLBCL 41.8% and MCL 11.2%. BMB was performed in 85.2%, CFM in 95.5% and PET/CT in 83.9%. A total of 123 patients (91.8%) received treatment: 21 (15.7%) rituximab monotherapy, 77 (57.5%) rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP), 15 (11.2%) R plus other chemotherapy combinations, 2 (1.5%) radiotherapy (RT), and 8 (5.9%) other treatments. The AA-baseline stage was I: 18.0%, II: 15.0%, III: 28.6%, and IV: 38.35%. The addition of BMB, FCM, and PET/CT resulted in upstaging in 16.5% (AA-BMB), 16.3% (AA-FCM), and 7.3% (AA-PET) of the patients, respectively. Interestingly, upstaging by BMB and FCM was higher in aggressive NHL (21.4% and 23.5%, respectively) compared to indolent NHL (11.1% and 8.2%, respectively). This fact was also observed in the case of PET/CT, although with a lower percentage of upstaging compared to the other two techniques (10.0% in aggressive NHL and 4.1% in indolent NHL). The 5-year PFS and OS were 63.5% (95%CI 53.1-76.0) and 79.3% (95%CI 70.9-88.8), respectively. Differences in discrimination for PFS and OS were observed through the 3 techniques, but varied according to NHL group. In indolent NHL, the addition of BMB findings (AA-BMB) to AA-baseline showed a higher C-Index for PFS (0.620 vs. 0.591) and for OS (0.683 vs. 0.597). In the AIC analysis, AA-PET showed the highest discrimination for PFS (86.2941 vs. 111.7166 for the AA-baseline). However, according to the CPE analysis, AA-BMB showed the highest discrimination for PFS compared with AA-baseline (0.604 vs. 0.6038, respectively) and, in particular, for OS (0.7337 vs. 0.5654, respectively). A similar trend was observed in aggressive NHL, with a C-Index for PFS showing identical results (0.572) for both AA-BMB and AA-PET, without clear predominance of one test over the others. Conclusion In the staging of NHL, evaluation of bone marrow involvement through BMB or high sensitivity FCM upstages patients in a higher proportion compared to PET/CT. Including BMB in the staging process results in a slight improvement in survival discrimination, particularly in indolent NHL. Our data are in favor of keeping assessment of BM through BMB in the work-up at diagnosis, at least in indolent NHL. * JA. S. and E. R. are equally contributed authors.