LAUSR

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma (NHL). Although introduction of rituximab in the upfront treatment of DLBCL was followed by a major improvement in response and survival, up to 40% of patients are expected to relapse while 10% of patients are present with refractory disease, demonstrating an unmet need for novel prognostic factors. The presence of monoclonal gammopathy (MG) has been a well-described phenomenon in a wide variety of lymphoproliferative disorders, especially in the context of differentiated indolent lymphomas; however, regarding DLBCL, both the frequency and the prognostic significance of MG remains equivocal. The aim of this study was to investigate the incidence and prognostic value of MG in a consecutive series of DLBCL patients treated in the rituximab era. This study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of University of Athens, Greece. Out of 211 DLBCL patients, an immunofixation assay at diagnosis was available for 138 patients who were therefore included in the analyses. Correlation of baseline demographics clinical characteristics among the study population as well as patients with MG was assessed using the paired chi-square test; results are shown in Table 1. MG was documented in 19 patients (13.8%); (IgM in 11 patients, IgG in four patients, both IgG and IgA in two, and monoclonal free light chains in two patients). Patients with MG were more likely to be diagnosed in a more advanced stage, had B symptoms and poorer PS, and were more prone to be assigned in the intermediate-high/high risk groups of IPI. MUM-1 was positive in the vast majority of patients with MG (17 out of 18 patients). During a mean follow-up of 41.1 months, 23 patients relapsed (23.5%) and 38 died (27.5%). Complete response (CR/CRu) was documented in 77.6% of patients. Patients with MG were less likely to reach CR: 8/17 MG patients demonstrated CR/CRu (P = .003). Notably, non-IgM patients exhibited a more aggressive disease course, as none of them achieved CR, and all of them had died at the end of follow-up. Survival curves were plotted using the KaplanMeier method and compared using the log-rank test. Cox proportional hazards regression was used for univariate and multivariate analysis of overall survival (OS) and progression-free survival (PFS). Presence of MG was associated with significantly inferior PFS (P = .003) (Figure 1A). Regarding paraprotein isotype, IgG/IgA was associated with shorter PFS (P < .001), whereas such an association was not demonstrated for IgM (P = .159) (Figure 1B). In a univariate Cox regression analysis, presence of any MG yielded a hazard ratio (HR) of 2.59, 95% confidence intervals (CI), 1.34-5.01. Regarding the isotypes, IgG/IgA correlated strongly with inferior PFS, whereas no significant association with PFS was found for IgM (Table S1). The potential independent association of MG with PFS was assessed in a multivariate model, incorporating IPI, and at first, presence of any MG and thereafter presence of either IgM or IgG/IgA. In both models, high IPI was associated with inferior PFS, as expected, and presence of any MG was marginally associated with PFS. Presence of IgM paraprotein remained insignificant, whereas presence of IgG/IgA remained significantly associated with PFS (Table S2). Similarly, Kaplan-Meier survival analysis demonstrated a significant association between MG and inferior OS (P = .004) (Figure 1C). Regarding the isotype, IgG/IgA paraprotein correlated strongly with inferior OS (P < .001), whereas no significant association was demonstrated for IgM (P = .435) (Figure 1D). In a univariate Cox regression model, MG was associated with inferior OS. Presence of MG with an IgM isotype was not associated with OS, whereas presence of IgG/IgA isotype correlated strongly with OS (Table S1). In a multivariate model incorporating IPI as a confounding factor, only IgG/IgA isotype was independently associated with inferior OS (Table S2). In our cases series, MG was documented in 13.8% of DLBCL patients. Economopoulos et al in a previous single-institution Greek study has reported the presence of MG in 14.6% DLBCL patients, whereas other studies report comparable rates. Regarding the specific isotypes, IgM has been the most prevalent isotype in several cases series; albeit Kim et al and Li et al demonstrated a predisposition towards IgG isotype. Regarding the prognostic significance of MG in DLBCL, there has been limited yet concordant data indicating a dismal prognosis for DLBCL patients with MG. However, there has been dispute regarding the role of specific isotypes in prognosis. Indeed Zhang et al have suggested that presence of MG regardless of isotype is associated with a detrimental effect on OS and PFS. On the other hand, Li et al suggested that presence of non-IgM MG was a negative prognostic factor for both OS and PFS, whereas Cox et al suggest that IgM-secretion harbors a negative effect on both event-free survival (EFS) and OS. In our study, presence of non-IgM MG was recognized as an independent negative prognostic factor for both PFS and OS. Received: 16 July 2019 Revised: 20 September 2019 Accepted: 21 September 2019 DOI: 10.1002/hon.2685