Diffuse Large B-cell Lymphoma : Treatment and Support

Diffuse large B-cell lymphoma (DLBCL) is a neoplasm of morphologically medium to large B-lymphoid cells. The most recent WHO classification of tumors of hematopoietic and lymphoid tissues (Swerdlow et al. 2017) divides DLBCL into DLBCL, NOS with distinct morphological (centroblastic, immunoblastic, anaplastic, rare) and molecular (germinal center B-cell, activated B-cell) subtypes, other lymphomas of large B cells, high-grade B-cell lymphoma (with MYC and BCL2 and/or BCL6 rearrangements or NOS), and B-cell lymphoma, unclassifiable. With some important exceptions, diagnostic work-up and treatment are identical in all DLBCL subtypes. It is beyond the scope of this article to fully describe the exceptions; we mention the most important differences but otherwise focus on transplantation for patients with relapsed/ refractory DLBCL. DLBCL is the most frequent lymphoma subtype and accounts for approximately one third of newly diagnosed lymphoma cases worldwide. In Europe, the 10-year prevalence of DLBCL is estimated at 43.3 per 100,000 per year (Smith et al. 2015); the age-adjusted incidence rate of DLBCL reported by the US Surveillance, Epidemiology, and End Results (SEER) program is 7.14/100,000 person-years (Howlader et al. 2017). The disease is slightly more frequent in men than in women; it mostly is a disease of the elderly (median patient age beyond 60 years) but can occur also in children and adolescents.


Definition and Epidemiology
Diffuse large B-cell lymphoma (DLBCL) is a neoplasm of morphologically medium to large B-lymphoid cells. The most recent WHO classification of tumors of hematopoietic and lymphoid tissues (Swerdlow et al. 2017) divides DLBCL into DLBCL, NOS with distinct morphological (centroblastic, immunoblastic, anaplastic, rare) and molecular (germinal center B-cell, activated B-cell) subtypes, other lymphomas of large B cells, high-grade B-cell lymphoma (with MYC and BCL2 and/or BCL6 rearrangements or NOS), and B-cell lymphoma, unclassifiable.
With some important exceptions, diagnostic work-up and treatment are identical in all DLBCL subtypes. It is beyond the scope of this article to fully describe the exceptions; we mention the most important differences but otherwise focus on transplantation for patients with relapsed/ refractory DLBCL.
DLBCL is the most frequent lymphoma subtype and accounts for approximately one third of newly diagnosed lymphoma cases worldwide. In Europe, the 10-year prevalence of DLBCL is estimated at 43.3 per 100,000 per year (Smith et al. 2015); the age-adjusted incidence rate of DLBCL reported by the US Surveillance, Epidemiology, and End Results (SEER) program is 7.14/100,000 person-years (Howlader et al. 2017). The disease is slightly more frequent in men than in women; it mostly is a disease of the elderly (median patient age beyond 60 years) but can occur also in children and adolescents.

Diagnosis
The diagnosis is made according to the WHO classification from a sufficiently large surgical specimen or excisional lymph node biopsy; needle biopsies are not recommended. Beyond morphological evaluation by an experienced pathologist, determination of the immunophenotype of the malignant cells (positivity of malignant cells for CD19 and CD20 must be documented because of its therapeutic consequences) and determination of the cell of origin by adequate molecular methods are required (Swerdlow et al. 2017).

Classification
The large B-cell lymphomas comprise the morphological and molecular subtypes of DLBCL, NOS, high-grade B-cell lymphomas, and B-cell lymphomas, unclassifiable (see paragraph 1).
The WHO classification describes 12 other (and 1 provisional) lymphomas of large B cells. Among these, primary diffuse large B-cell lymphoma of the CNS (PCNSL), lymphomatoid granulomatosis (LG), primary mediastinal large B-cell lymphoma (PMBCL), and plasmablastic lymphoma not only show significant differences in pathogenesis and clinical manifestation but in most centers are treated different from classical DLBCL. Exact subtyping of diagnostic specimens taking into account their origin (e.g., primary cutaneous DLBCL, leg type; PCNSL; primary effusion lymphoma) is important. In order to fulfill all WHO requirements, the cell of origin (GCB-or ABC-subtype by IHC or gene expression profiling) and the presence/absence of distinct chromosomal translocations (BCL6, BCL2, MYC by FISH testing or IHC) must be determined.

Risk Factors
The International Prognostic Index (IPI) remains the most important tool in order to estimate the prognosis of patients with DLBCL (Ziepert et al. 2010). The IPI takes into account five factors (age, stage, LDH, performance status, and number of extranodal sites involved). Patients within the low (IPI 0, 1), low-intermediate (IPI 2), highintermediate (IPI 3), and high-risk group (IPI 4,5) can expect 3-year overall survival of 91.4%, 80.9%, 65.1%, and 59.0%, respectively, if treated with R-CHOP or one of its variants. Other clinical risk factor models (R-IPI; NCCN-IPI) (Sehn et al. 2007;Zhou et al. 2014, respectively) have been proposed, but advantages over the original IPI seem limited.

First-Line Treatment
First-line treatment of patients with DLBCL generally consists of RTX at standard dose (375 mg/ sqm) in combination with CHOP (CY, DOXO, VCR, PRD) or one of its variants such as ACVBP, CHOEP, or DA-EPOCH chemotherapy. Six cycles of R-CHOP are generally used. However, the cycle number can be reduced to four without jeopardizing treatment outcome in patients with IPI 0. Patients with early disease (IPI 0 and 1) have been treated with abbreviated chemotherapy and involved-field radiotherapy (RT). Recent studies do not support a role for RT in such patients (Lamy et al. 2018). In patients with IPI 2-5, radiotherapy to bulky and extranodal disease is regularly recommended after R-CHOP in some but not in the majority of countries.
Several studies evaluated the role of consolidative high-dose therapy followed by auto-HSCT in the RTX era. The French (Gouill et al. 2007), Italian (Chiappella et al. 2017), and German ) studies failed to demonstrate an advantage of auto-HSCT over conventional chemotherapy. The only American study (Stiff et al. 2013) reported an advantage of auto-HSCT in younger patients with high-risk disease (age-adjusted IPI 3); however, this study included patients treated with CHOP only and patients with T-cell lymphoma and as a consequence was underpowered in order to show a significant advantage of auto-HSCT over R-CHOP (Schmitz et al. 2014). In young patients who remain PET positive after two cycles of chemo-immunotherapy, auto-HSCT is performed in some but not in the majority of countries.
Patients with PCNSL or DLBCL with primary involvement of testicles must receive chemotherapy penetrating into the CNS.
More aggressive chemotherapies (CHOEP, DA-EPOCH, or ACVBP) in combination with RTX with or without RT are recommended in patients with PMBCL or plasmablastic lymphoma and in patients with high-intermediate and high-risk disease (Ghielmini et al. 2013;Recher et al. 2011). If patients with DLBCL of ABC subtype or rearrangement of MYC and BCL2 and/or BCL6 (double-or triple-hit lymphoma) should receive more aggressive therapy, therapy remains controversial (Friedberg 2017;Staiger et al. 2017). The same holds true for the value of targeted therapies (ibrutinib, lenalidomide, and others) given in addition to R-CHOP. Prospective randomized studies addressing these questions have been closed but not published yet.

Second-Line Treatment
The principles of management of relapsed and refractory DLBCL are shown in Table 85.1. All chemotherapy-based salvage regimens cause hematologic toxicity in many cases necessitating RBC and platelet transfusions. Mucositis, gastrointestinal toxicities, neutropenic fever, and infections are reported in a significant proportion of patients. Nephrotoxicity, hepatotoxicity, and other non-hematologic toxicities are also observed. Failure to mobilize hematopoietic stem cells in 10-20% of cases occurs with all salvage regimens. Efficacy of different salvage options is shown in Table 85.2.

Autologous HSCT
Autologous HSCT is still considered the standard treatment for patients with refractory or relapsed (R/R) DLBCL. In the RTX era, however, the results of salvage therapy followed by auto-HSCT are less convincing than before, and the benefit of auto-HSCT even for those patients achieving PR or CR with salvage chemotherapy and RTX is limited (Crump et al. 2014(Crump et al. , 2017. In particular, patients with refractory disease or early relapse pretreated with RTX as part of first-line therapy rarely achieve long-term remissions after auto-HSCT. In the CORAL study, 3-year PFS for such patients was only 23% although those proceeding to auto-HSCT showed 3-year PFS of 39%. Alternative treatment, e.g., allo-HSCT, may be more adequate. EBMT indications (Sureda et al. 2015) for auto-HSCT in DLBCL are shown in Table 85.3. Auto-HSCT is generally not recommended as part of first-line therapy in DLBCL although recent data on PET-guided auto-HSCT are promising (Casasnovas et al. 2017). We discourage auto-HSCT for patients with refractory disease not responding to salvage therapy.  (Thieblemont et al. 2011). Therefore, salvage strategies should take into account individual patient characteristics (age and comorbidities) considering potential cumulative hematologic and non-hematologic toxicity and the possibility of harvesting stem cells. Cardiac, pulmonary, renal, and liver function should be evaluated prior to treatment -The objective of salvage chemotherapy is to induce a complete or partial response indicating that the tumor remains chemosensitive, this having a major impact on outcome after transplantation. PET negativity after salvage therapy is a surrogate marker of chemosensitivity and predicts patient outcome after auto-HSCT

HSC Source
PBSC is used in >90% of auto-HSCT.
Other high-dose regimens have been used sometimes because of shortage of MEL or BCNU. Recent publications suggest that the BEAC (CY) and TEAM (TT) regimens show efficacy and toxicity similar to BEAM in most if not all lymphoma subtypes (Robinson et al. 2018;Sellner et al. 2016).

Prognostic Factors
Adverse prognostic factors for auto-HSCT identified in prospective studies include early relapse within 12 months of induction therapy, prior exposure to R, secondary age-adjusted IPI, poor performance status, and involvement of two or more extranodal sites at relapse.

Consolidation Treatment After Auto-HSCT
There are no data and no recommendation for consolidative therapy after auto-HSCT for DLBCL. In the CORAL study Gisselbrecht et al. (2012), RTX maintenance did not improve outcome.

Tandem Transplantation
No data from the RTX era are available.

Relapse After Auto-HSCT
Patients relapsing after auto-HSCT generally have a poor prognosis. Therapeutic options are limited. Data on new drugs specifically used post auto-HSCT are not available. Results of allo-HSCT after failure of auto-HSCT are reported below. CAR T cells may be a therapeutic option; however, no more than anecdotal data are available. The only prospective randomized clinical trial reported so far (Glass et al. 2014) compared GVHD prophylaxis including RTX or not after MAC and allo-HSCT for R/R DLBCL. No significant difference between patients receiving or not receiving RTX in addition to standard MMF and TAC for GVHD prophylaxis was found. OS was 52% for all 82 patients randomized with significant differences between patients transplanted from a MUD or MMUD and patients receiving or not receiving ATG. For patients transplanted from matched family donors or MUD receiving ATG, OS was 64.7%. These data show that allo-HSCT is a valid alternative to any other treatment for patients relapsing after failure from auto-HSCT. Allo-HSCT should also be considered for patients with early relapse after first-line R-CHOP or similar.

Stem Cell Source
PBSC is the preferred stem cell source for allo-HSCT. The use of haploidentical donors has somewhat increased the use of BM in some of the series.

Donor Selection
In recent years, there has been a significant increase in the use of haploidentical donors for allo-HSCT after the introduction of PT-CY.
Retrospective analyses from EBMT and CIBMTR (Kanate et al. 2016;Ghosh et al. 2016) suggest that allo-HSCT from HLA-identical family and URD or from haploidentical donors give comparable results. However, no prospective clinical trials comparing haploidentical donors versus HLA-identical siblings and MUD have been published so far.

Conditioning
RIC regimens reduce NRM after transplantation in many indications but also tend to increase RI after transplantation. Because no prospective clinical trials demonstrating the superiority of one conditioning regimen over another have been reported, the question if RIC or MAC should be preferred cannot generally be answered. Aggressive disease not completely responding to salvage therapy and high tumor are situations where MAC should strongly be considered.

Prognostic Factors
The most important adverse prognostic factor that impacts long-term outcome of patients being treated with allo-HSCT is disease status before the treatment. However, unlike the situation with auto-HSCT, also patients not perfectly responding to salvage therapy, e.g., patients with minor response or stable disease, may benefit from allo-HSCT.

The Use of Allo-HSCT in the Era of New Drugs and CAR T Cells
In contrast to the situation for other lymphomas, e.g., Hodgkin's disease, new drugs have not really affected the role and positioning of allo-HSCT in patients relapsing/progressing after auto-HSCT. Lenalidomides, ibrutinib, polatuzumab, or checkpoint inhibitors with or without chemotherapy can be used to bring more patients to transplantation. None of these drugs can substitute for allo-HSCT because remissions are mostly transient, and no cures have been achieved. CAR T cells (Schuster et al. 2017;Neelapu et al. 2018) 1999-2002 2003-2006 2007-2010 2011-2014 p = 0.11 1990-1998 Fig. 85.2 Long-term outcomes of auto-and allo-HSCT in patients with R/R DLBCL (EBMT data base, with permission) selection of patients (e.g., including PET imaging for patient selection), and improved salvage strategies results. • Allo-HSCT is the only curative treatment option for patients with refractory disease and those relapsing after auto-HSCT. Patients with early relapse (< 12 months after first-line treatment) should be considered for allo-HSCT. Conditioning should be guided by the individual clinical situation. Haploidentical transplants may substitute for unrelated donor transplants in the near future. New drugs have not really changed the treatment algorithm for DLBCL. The role of CAR T cells is under study.