INTRODUCTION — Diffuse large B cell lymphoma (DLBCL) is an aggressive non-Hodgkin lymphoma (NHL) and it is the most common histologic category of NHL. Initial treatment of DLBCL is dependent upon the extent of disease (table 1). For treatment purposes, patients with DLBCL are classified as having either limited stage disease (usually Ann Arbor stage I or II) or advanced stage (stage III or IV); some patients with stage II DLBCL with bulky disease (ie, >10 cm) or constitutional symptoms are treated for advanced disease.

The initial treatment of adults with advanced stage DLBCL is discussed here.

Diagnosis, staging, and prognostic features of DLBCL and treatment of limited stage DLBCL and relapsed disease are presented separately.

●(See "Diffuse large B cell lymphoma (DLBCL): Suspected first relapse or refractory disease in medically-fit patients".)

●(See "Pretreatment evaluation, staging, and response assessment of non-Hodgkin lymphoma".)

●(See "Epidemiology, clinical manifestations, pathologic features, and diagnosis of diffuse large B cell lymphoma".)

●(See "Clinical presentation and initial evaluation of non-Hodgkin lymphoma".)

SPECIAL CONSIDERATIONS DURING THE COVID-19 PANDEMIC — The coronavirus disease 2019 (COVID-19) pandemic has increased the complexity of cancer care. Important issues include balancing the risk from treatment delay versus harm from COVID-19, ways to minimize negative impacts of social distancing during care delivery, and appropriately and fairly allocating limited health care resources. These issues and recommendations for cancer care during the COVID-19 pandemic are discussed separately.

●(See "COVID-19: Considerations in patients with cancer".)

PRETREATMENT EVALUATION — To best treat patients with non-Hodgkin lymphoma (NHL), the initial evaluation must establish the precise histologic subtype, the extent and sites of disease, and the performance status of the patient. General approaches to the diagnostic work-up and staging of NHL are presented separately (table 1). (See "Clinical presentation and initial evaluation of non-Hodgkin lymphoma" and "Pretreatment evaluation, staging, and response assessment of non-Hodgkin lymphoma".)

Clinical assessment — The pretreatment evaluation both determines the extent of the disease and provides information about the individual's comorbidities that are likely to have an impact on treatment options. In addition to a history and physical examination, it is our practice to perform the following pretreatment studies in patients with DLBCL:

●Laboratory studies include a complete blood count with differential, chemistries with liver and renal function and electrolytes, lactate dehydrogenase (LDH), albumin, hepatitis B, and uric acid. All patients with DLBCL should be encouraged to undergo serologic testing for human immunodeficiency virus (HIV). (See "Hepatitis B virus reactivation associated with immunosuppressive therapy".)

●Unilateral bone marrow biopsy is suggested for most patients. Although positron emission tomography (PET) has a high sensitivity and specificity for detecting marrow involvement in patients with DLBCL, it may miss low volume, diffuse involvement [1]. In addition, only pathologic review of a bone marrow biopsy can determine whether such involvement is concordant or discordant, a distinction that has prognostic value [2]. As such, bone marrow biopsy is an important part of the evaluation of the patient without evidence of bone involvement by PET. In contrast, bone marrow biopsy may not necessary if the PET indicates bone or marrow involvement [3]. (See "Pretreatment evaluation, staging, and response assessment of non-Hodgkin lymphoma", section on 'Bone marrow examination'.)  

●Patients with DLBCL do not require routine pretreatment lumbar puncture (LP) for evaluation of cerebrospinal fluid. Some experts suggest performing an LP for patients with neurologic findings on clinical evaluation; evidence of leptomeningeal involvement by magnetic resonance imaging (MRI; or computed tomography [CT]); >1 extranodal site; or involvement of certain extranodal sites (eg, kidney, adrenal gland, testis). Evaluation and management of patients with CNS involvement is discussed separately. (See "Clinical presentation and diagnosis of secondary central nervous system lymphoma".)

●Echocardiogram or radionuclide ventriculography (MUGA scan) should be performed to assess left ventricular ejection fraction.

●Men and women with childbearing potential should receive counseling about the potential effect of treatment on their fertility and options for fertility-preserving measures. Given the urgent need for treatment with DLBCL, options for women are limited, but men can often participate in sperm banking. (See "Overview of fertility and reproductive hormone preservation prior to gonadotoxic therapy or surgery".)

●Integrated whole body PET/CT is the preferred pretreatment imaging study. DLBCL is considered a typically FDG-avid tumor such that involved areas should show activity on PET scan. A dedicated contrast-enhanced CT scan may be required in addition to the PET/CT to define the extent of disease in special situations, such as in the setting of lymphadenopathy close to bowel or if there is compression or thrombosis of blood vessels. (See "Pretreatment evaluation, staging, and response assessment of non-Hodgkin lymphoma", section on 'Routine imaging studies'.)

Identify disease subtype — Once the diagnosis of DLBCL is confirmed, the next step includes an evaluation for those abnormalities that help predict outcome following treatment. Prognosis in DLBCL depends on several variables: host factors (age, performance status, comorbidities), stage, response to therapy, and disease biology. Disease biology is dictated by genetic and molecular events. (See "Prognosis of diffuse large B cell lymphoma", section on 'Molecular genetics'.)

We suggest that the following tests be performed to assess molecular risk in all newly diagnosed patients:

●Evaluation of cell of origin (gene expression profile [GEP] analysis, immunohistochemistry algorithms, or LymphCx platform) – GEP is the gold standard for determining cell of origin, but it requires fresh tissue and is not routinely available in practice. Given their ease of use, immunohistochemistry algorithms (eg, Tally and Hans) (figure 1) are the most commonly used method; concordance with GEP is approximately 80 percent. While not yet widely available, LymphCx can be performed on formalin-fixed paraffin-embedded tissue and is highly concordant with GEP. When available, LymphCx may replace immunohistochemistry algorithms for cell of origin assignment in DLBCL.

Using this information, an individual case may be subclassified according to the cell of origin as one of the following:

•Germinal center B cell (GCB) DLBCL – Cases with GCB DLBCL identified by GEP, immunohistochemistry algorithms, or LymphCx without double hit DLBCL. These patients have a relatively good prognosis following standard therapy with R-CHOP. (See 'Germinal center B cell type DLBCL' below.)

•Activated B cell (ABC) DLBCL or non-GCB DLBCL – Cases with non-GCB DLBCL identified by GEP, immunohistochemistry algorithms, or LymphCx without double hit DLBCL have high relapse rates and a less favorable prognosis following treatment with R-CHOP. Patients with non-GCB DLBCL should be referred for trials investigating novel therapies.  

●Evaluation of MYC, BCL2, and BCL6 status (fluorescence in situ hybridization [FISH] or immunohistochemistry) – FISH is preferred in all cases. An acceptable alternative is to perform immunohistochemistry for MYC and BCL2. Cases with MYC overexpression on immunohistochemistry should have further testing for MYC, BCL2, and BCL6 gene rearrangements by FISH.

This information is used to determine whether the lymphoma should be classified as:

•Double hit DLBCL – MYC translocation plus gene rearrangement of BCL2, BCL6, or both; double hit genetic events can be found in both GCB and ABC types of DLBCL. Patients with double hit DLBCL are candidates for more aggressive chemotherapy regimens (eg, da-EPOCH-R) given high rates of relapse and poor survival following treatment with standard R-CHOP. (See 'Double hit or double expressor DLBCL' below.)

Primary large B cell lymphoma of the mediastinum (PMBL) is a distinct clinicopathologic entity that must be distinguished from DLBCL. PMBL is an aggressive tumor arising in the mediastinum from the thymic (medullary) B cell. Patients present with a locally invasive anterior mediastinal mass that frequently extends into local structures, resulting in superior vena cava syndrome, effusions, and symptoms of cough, dyspnea, dysphagia, and/or hoarseness. Management differs from DLBCL. The diagnosis and management of PMBL is discussed separately. (See "Primary mediastinal large B cell lymphoma".)

Assess need for CNS prophylaxis — Patients with DLBCL have a risk of central nervous system (CNS) recurrence of approximately 5 percent, and the decision to administer CNS prophylaxis should be decided on a case-by-case basis. The following populations have a particularly high risk of CNS recurrence, which may justify CNS prophylaxis:

●An elevated serum lactate dehydrogenase (LDH) level and involvement of >1 extranodal site

●Four or more of the following risk factors: elevated serum LDH, serum albumin <35 g/L, >60 years of age, retroperitoneal lymph node involvement and involvement of >1 extranodal site

●Lymphomatous involvement of certain organs (eg, testes)

●Concurrent bone marrow involvement

The decision to administer CNS prophylaxis and the choice of regimen is discussed in more detail separately. (See "Initial treatment of limited stage diffuse large B cell lymphoma", section on 'CNS prophylaxis'.)

INITIAL TREATMENT — Advanced stage DLBCL is that which cannot be contained within one irradiation field (usually Ann Arbor stage III or IV disease). This population accounts for 60 to 70 percent of patients with DLBCL. Without therapy, patients with DLBCL have survival times measured in months. Standard treatment with combination chemotherapy plus rituximab can cure a percentage of patients resulting in survival rates of approximately 70, 60, and 45 percent at 3, 5, and 10 years after diagnosis.

Advanced stage DLBCL is treated primarily with systemic chemotherapy plus immunotherapy with the recombinant anti-CD20 antibody rituximab. The preferred chemotherapy depends in part on the molecular subtype. While general treatment principles for advanced stage DLBCL apply to most disease presentations and patient populations, involvement of certain organs, particularly when occurring as primary sites of disease, has unique therapeutic implications. In addition, older adults and patients with cardiac disease or human immunodeficiency virus (HIV) infection require special consideration. In addition, there are some special scenarios (eg, testicular lymphoma) in which central nervous system prophylaxis is required. These are discussed in detail separately. (See 'Special scenarios' below and "Initial treatment of limited stage diffuse large B cell lymphoma", section on 'CNS prophylaxis'.)

Targeting CD20 — There is broad consensus that initial treatment of all patients with DLBCL should include rituximab (or another anti-CD20 monoclonal antibody), based on randomized clinical trials that demonstrated superior survival with little additional toxicity for chemotherapy plus rituximab compared with the same chemotherapy regimen alone.

The survival benefit of including rituximab with chemotherapy has been observed in adults of all ages:

●Three large randomized trials, which included approximately 2000 patients over age 65 with advanced stage DLBCL, reported significantly higher overall survival (OS) at three (approximately 70 versus 57 percent for CHOP alone), five (58 versus 45 percent), and 10 (44 versus 28 percent) years [4-8]. There was no difference in the rates of severe or serious side effects after CHOP with or without rituximab.

●A clinical trial that included 824 patients <60 years of age with DLBCL (28 percent stage III/IV and 48 percent with bulky disease) reported superior event-free survival (EFS) at three (79 versus 59 percent, respectively) and five years (74 versus 56 percent), and superior OS at three (93 versus 84 percent) and five years (90 versus 80 percent) [9,10].

Rituximab is generally administered intravenously at a dose of 375 mg/m² on day 1 of each chemotherapy cycle; a phase 3 trial reported no benefit associated with dose intensification of rituximab during the first four cycles of R-CHOP therapy [11]. The major toxicities of rituximab include infusion reactions, infections related to immunosuppression, reactivation of hepatitis B in patients who are positive for hepatitis B surface antigen (HBsAg) or antibodies against hepatitis B core antigen (anti-HBc), and, rarely, progressive multifocal leukoencephalopathy caused by reactivation of JC virus infection. (See "Infusion-related reactions to therapeutic monoclonal antibodies used for cancer therapy" and "Secondary immunodeficiency induced by biologic therapies", section on 'Rituximab' and "Hepatitis B virus reactivation associated with immunosuppressive therapy" and "Progressive multifocal leukoencephalopathy (PML): Epidemiology, clinical manifestations, and diagnosis".)

A subcutaneous formulation (rituximab-hyaluronidase) that uses a fixed dose and a shorter administration time is an acceptable alternative for patients who have tolerated at least one full dose of intravenous rituximab [12]. Randomized trials have demonstrated comparable efficacy and safety of the two formulations in patients with DLBCL and other CD20-expressing B cell neoplasms [4-9,13-16].

Molecular subtype-specific therapy — R-CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone plus rituximab) (table 2) has been the standard chemotherapy regimen used for the treatment of most patients with DLBCL. This preference is based on numerous randomized trials conducted prior to the advent of rituximab-containing regimens, which compared CHOP with many other anthracycline-based regimens. These regimens offered no improvement in remission rate, disease-free survival (DFS), or OS and were associated with increased toxicity [17-21].

Subsequent studies have evaluated the efficacy of R-CHOP in different genetic and molecular subtypes of DLBCL and suggest its efficacy differs among these populations. As an example, an analysis of 344 patients with de novo DLBCL (51 percent with advanced stage disease) treated with curative intent with R-CHOP reported the following estimated outcomes at five years according to molecular subtype [22]:

●Germinal center B cell (GCB) type DLBCL – Freedom from progression (FFP) 76 percent; progression-free survival (PFS) 73 percent; disease-specific survival (DSS) 82 percent; OS 78 percent.

●Activated B cell (ABC) type DLBCL – FFP 51 percent; PFS 48 percent; DSS 61 percent; OS 56 percent.

●Double expressor DLBCL (MYC expression and BCL2 expression by immunohistochemistry) – FFP 51 percent; PFS 48 percent; DSS 58 percent; OS 54 percent.

These results clearly demonstrate that ABC and double protein DLBCL are associated with inferior outcomes after R-CHOP. Other studies have shown even worse outcomes for patients with double hit DLBCL (MYC translocation plus gene rearrangement of BCL2, BCL6, or both) [23-25]. While there is still room for improvement, R-CHOP remains an effective therapy for patients with GCB type DLBCL. In contrast, those with ABC type DLBCL or double hit DLBCL have unacceptably high rates of relapse and poor survival following treatment with R-CHOP.

●For most patients with advanced stage GCB DLBCL, we offer six cycles of R-CHOP-21 (cyclophosphamide, doxorubicin, vincristine, and prednisone plus rituximab, given every 21 days) (table 2). We cap the total vincristine dose at 2 mg to limit neurotoxicity; some experts instead cap the vincristine dose at 3 mg [26]. (See 'Germinal center B cell type DLBCL' below.)

●For patients with advanced stage ABC type DLBCL, we encourage enrollment on a clinical trial evaluating the incorporation of novel agents (eg, R-CHOP plus lenalidomide; R-CHOP plus ibrutinib; R-CHOP plus bortezomib). The more intensive R-ACVBP regimen (rituximab, doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone, followed by consolidation with methotrexate and leucovorin) is an acceptable alternative in settings where vindesine is available. (See 'Activated B cell type DLBCL' below.)

●For patients with advanced stage double hit DLBCL, we encourage enrollment on a clinical trial. For those treated off study, we suggest six to eight cycles of dose-adjusted (da)-EPOCH-R (etoposide, doxorubicin, vincristine, cyclophosphamide, and prednisone plus rituximab) (table 3) with a goal of administering two cycles beyond attainment of a complete remission (CR). (See 'Double hit or double expressor DLBCL' below.)

●For the larger group of patients with advanced stage double expressor DLBCL, we encourage enrollment on a clinical trial. For those treated off study, R-CHOP is most appropriate given the paucity of data regarding other regimens in this population. (See 'Double hit or double expressor DLBCL' below.)

Support for these preferences is provided in the sections below.

Adjunctive care — Supportive care is a critical component to the treatment of patients with lymphoma. Toxicity can result from the chemotherapy or from the rapid elimination of a large tumor burden (ie, tumor lysis syndrome). Life-threatening adverse effects of therapy include tumor lysis syndrome, infection, and anaphylaxis. Other acute side effects include mucositis, nausea and vomiting, alopecia, and damage to the liver, heart, nerves, kidneys, and lungs. A few aspects of adjunctive care require specific mention:

●Infection risk – Both R-CHOP-21 and da-EPOCH-R result in neutropenia and carry a risk of febrile neutropenia. All patients receiving da-EPOCH-R should receive primary prophylaxis with hematopoietic growth factors. For those receiving R-CHOP-21, prophylactic growth factors are considered on an individual basis, particularly for high-risk patients such as those with pre-existing neutropenia, advanced disease, poor performance status, and patients age 65 years or older. Patients who develop neutropenic fever require prompt evaluation and immediate administration of parenteral antibiotics tailored to the prominent organisms and resistance patterns of the institution. (See "Use of granulocyte colony stimulating factors in adult patients with chemotherapy-induced neutropenia and conditions other than acute leukemia, myelodysplastic syndrome, and hematopoietic cell transplantation", section on 'Primary prophylaxis' and "Treatment of neutropenic fever syndromes in adults with hematologic malignancies and hematopoietic cell transplant recipients (high-risk patients)".)

●Emesis risk – R-CHOP-21 and da-EPOCH-R chemotherapy are highly emetogenic and prophylactic antiemetic therapy is indicated. This is discussed in more detail separately. (See "Prevention and treatment of chemotherapy-induced nausea and vomiting in adults", section on 'Regimens with a high risk of delayed emesis'.)

●Tumor lysis syndrome – Tumor lysis syndrome is an oncologic emergency caused by massive tumor cell lysis and the release of large amounts of potassium, phosphate, and uric acid into the systemic circulation. Deposition of uric acid and/or calcium phosphate crystals in the renal tubules can result in acute kidney injury. Patients who have advanced stage disease are at high risk for developing tumor lysis syndrome with the initiation of treatment. The diagnosis, prevention, and treatment of tumor lysis syndrome are discussed in more detail separately. (See "Tumor lysis syndrome: Definition, pathogenesis, clinical manifestations, etiology and risk factors".)

●Infusion reactions – A small percentage of patients receiving their first infusion of rituximab will have severe infusion-related hypersensitivity reactions. It is standard practice to premedicate with acetaminophen (625 mg orally) and diphenhydramine (50 mg orally) before each infusion or at the time of the reaction; this practice attenuates but does not prevent these events. We often also give an H2 blocker, such as famotidine, 60 minutes prior to the rituximab infusion. With this practice, we rarely see reactions. This complication is discussed in more detail separately. (See "Infusion-related reactions to therapeutic monoclonal antibodies used for cancer therapy", section on 'Rituximab'.)

●Hepatitis reactivation – Rituximab therapy carries a risk of hepatitis B reactivation among patients positive for HBsAg or anti-HBc. The risk of hepatitis B reactivation is discussed in more detail separately. (See "Hepatitis B virus reactivation associated with immunosuppressive therapy".)

●Transfusions – Many patients will develop anemia and thrombocytopenia requiring transfusion support. In general, we transfuse packed red blood cells to all symptomatic patients with anemia or any asymptomatic patient with a hemoglobin ≤7 to 8 g/dL. We aim to maintain the hemoglobin between 8 and 9 g/dL, depending on the patient's age, symptoms, and comorbid conditions. We transfuse platelets prophylactically for patients with platelet counts <10,000/microL or to any patient with signs of overt bleeding, such as oral purpura. Blood products should be leukocyte depleted and irradiated to minimize the risk of immunologic-mediated events.

●Alopecia – Hair loss is a transient and usually (although not always) completely reversible consequence of chemotherapy that is often psychologically difficult. Patients should be informed of the likely side effect of hair loss. Options such as head wraps, hats, or wigs should be discussed in advance so that the patient can be more physically and emotionally prepared. (See "Alopecia related to systemic cancer therapy".)

Germinal center B cell type DLBCL

Choice of chemotherapy regimen — GCB DLBCL accounts for approximately 40 percent of de novo DLBCL overall [27].

For most patients with advanced stage GCB DLBCL, we suggest six cycles of R-CHOP-21 (cyclophosphamide, doxorubicin, vincristine, and prednisone plus rituximab, given every 21 days) (table 2). With this approach, estimated rates of PFS and OS at five years are 70 and 75 percent, respectively [22].

Compared to R-CHOP-21, no alternative regimen or variation in R-CHOP administration schedule has been shown to achieve superior outcomes, but some of the alternative regimens were more toxic. Informative studies include:

●da-EPOCH-R versus R-CHOP – A multicenter trial reported no differences in response rate or PFS or OS at two years among 491 patients randomly assigned to da-EPOCH-R versus R-CHOP [28]. Grade ≥3 adverse events were more common with da-EPOCH-R, including infection (17 versus 11 percent), febrile neutropenia (35 versus 18 percent), mucositis (8 versus 2 percent), and neuropathy (19 versus 3 percent).

●R-CHOP-14 versus R-CHOP-21 – A trial of 1080 patients with newly diagnosed DLBCL randomly assigned treatment with eight cycles of R-CHOP-21 versus six cycles of R-CHOP-14 followed by two additional doses of rituximab [29]. At a median follow-up of 46 months, there was no significant difference in OS or PFS between the two treatment groups. Another trial randomly assigned 600 older adults with DLBCL to eight cycles of R-CHOP-14 versus eight cycles of R-CHOP-21 and reported similar rates of CR, EFS, and OS with the two treatment regimens [30].

CHOP-14 has been associated with increased toxicity, including an increased risk of pneumocystic jirovecii pneumonia [31].

●R-CEOP – Some clinicians consider CEOP (which substitutes the epirubicin for doxorubicin) to have equivalent efficacy to CHOP, but this has not been proven. A trial that randomly assigned 217 patients with aggressive lymphoma to R-CEOP-14 versus R-CEOP-21 reported no significant difference in CR, overall response rate, or OS [32].

Number of treatment cycles — In the pre-rituximab era, the number of treatment cycles of CHOP given to patients with advanced disease has varied over the years from 6 to 12. In current practice, the number of treatment cycles given varies based upon whether the clinician chooses to administer a defined number of cycles independent of response (eg, administer six to eight cycles) or uses a response-based evaluation to determine the number of treatment cycles (eg, administer two cycles after the attainment of a CR).

Of the two options, we use a defined number of treatment cycles of chemotherapy. This is principally because the largest clinical trials that have investigated the use of chemotherapy for advanced DLBCL with or without rituximab have used this approach. The response-based evaluation relies upon an interim response as determined by computed tomography (CT) or combined positron emission tomography (PET)/CT imaging [33,34].

We suggest six cycles of R-CHOP-21 rather than eight. This preference is based on a desire to avoid unnecessary toxicity and an acceptance of the favorable outcomes seen following six cycles of R-CHOP-21 in the MiNT trial [9]. While no randomized trial has addressed the issue of cycle number using R-CHOP-21, one prospective, randomized trial has directly addressed this question with CHOP-14. The RICOVER-60 trial compared six and eight cycles of CHOP-14, each with and without the addition of rituximab in 1222 patients ages 61 to 80 years with aggressive non-Hodgkin lymphoma (80 percent DLBCL) [7]. Survival at three years was the same for those administered six versus eight cycles of R-CHOP-14 (78 and 73 percent).

Dose intensity — Delivery of full doses of chemotherapy is associated with improved clinical outcomes in DLBCL [35]. Dose reductions of CHOP chemotherapy are associated with lower survival rates, as highlighted in the following studies:

●In a retrospective analysis of 115 patients who were treated with anthracycline-based regimens, the amount of doxorubicin (ie, ≥75 percent of recommended dose) administered during the first 12 weeks of therapy was the single most important predictor of OS [36].

●A second retrospective review of 210 patients treated with CHOP-21 reported median survivals for patients who received >90 percent, 85 to 90 percent, and <85 percent of scheduled anthracycline of approximately 7, 3, and 1.7 years, respectively [37].

Is there a role for maintenance rituximab? — There is no well-defined role for maintenance therapy following initial treatment of advanced stage DLBCL. Both rituximab and lenalidomide can improve failure-free survival (FFS) following initial treatment of DLBCL, but neither agent improves OS when used as maintenance therapy.

We do not routinely offer maintenance therapy following R-CHOP for advanced DLBCL, because we judge that improved FFS, but not OS, does not outweigh the costs and toxicity of therapy.

Studies that have evaluated maintenance therapy for DLBCL include:

●A multicenter trial randomly assigned maintenance rituximab versus observation to 415 adults ≥60 years old who achieved CR after either R-CHOP or CHOP [6]. After three years, compared with observation, patients who received maintenance rituximab had improved FFS (53 versus 46 percent), but not OS; the FFS benefit of maintenance rituximab was limited to patients who previously received CHOP, but not R-CHOP. Longer-term follow-up, as reported in abstract form, also did not demonstrate an OS benefit of maintenance rituximab in this setting [38].

●A trial randomly assigned maintenance lenalidomide versus placebo to 650 older patients (60 to 80 years old) who achieved CR or partial response to R-CHOP therapy [39]. Lenalidomide resulted in superior PFS (HR, 0.7; 95% CI, 0.54 to 0.93) after three years follow-up, but did not improve OS after 52 months. Lenalidomide was associated with more grade 3/4 neutropenia (56 versus 22 percent, respectively) and cutaneous reactions (5 versus 1 percent) than placebo.

Is there a role for transplant? — High-dose chemotherapy followed by autologous hematopoietic cell transplantation (HCT) is not currently used in the initial treatment of DLBCL. This is principally because HCT is associated with significant morbidity and survival is the same in patients administered chemotherapy with or without HCT. This was best shown in a 2008 meta-analysis that included data from 15 randomized controlled trials with a total of 3079 patients treated for aggressive non-Hodgkin lymphoma [40]. Treatment-related mortality, EFS, and OS were the same whether patients were treated with conventional chemotherapy or high-dose chemotherapy followed by autologous HCT.

Randomized trials suggest that, even in high-risk disease, HCT at the time of relapse results in similar survival rates when compared with HCT in first CR:

●An intergroup trial (S9704) enrolled 397 patients with high-intermediate-risk and high-risk International Prognostic Index (IPI) scores according to age-adjusted index and, for those who attained an at least partial response after five cycles of CHOP-based therapy (CHOP or R-CHOP), randomized treatment to a total of six cycles of CHOP-based therapy followed by autologous HCT versus eight cycles of CHOP-based therapy alone [41]. Patients who relapsed after chemotherapy alone were allowed to proceed with autologous HCT as part of second-line therapy. The addition of autologous HCT did not improve OS at two years (74 versus 71 percent). In an unplanned subset analysis of 88 patients with high-risk IPI, autologous HCT resulted in superior two-year PFS (75 versus 41 percent) and OS (82 and 64 percent).

●An Italian trial assigned patients with high-intermediate or high IPI scores to treatment with R-CHOP every 14 days (122 patients; eight cycles) or rituximab plus high dose sequential (R-HDS) chemotherapy followed by autologous HCT (113 patients) [42]. There was no difference in CR rates (78 versus 76 percent, respectively), and after a median follow-up of five years, the three-year EFS was 62 versus 65 percent. Although DFS favored R-HDS after three years (79 versus 91 percent), this did not translate into improved OS (74 versus 77 percent), as there were higher rates of hematologic and infectious complications in the R-HDS arm.

●First-line autologous HCT showed no benefit over rituximab-based aggressive conventional chemotherapy (CHOP plus etoposide 300 mg/m² every two weeks) in a similarly designed trial conducted by the German High-Grade Non-Hodgkin Lymphoma Study Group (DSHNHL) [43].

●In another trial, 162 patients with aggressive B cell lymphoma and a high-intermediate-risk or high-risk age-adjusted IPI score were randomly assigned to R-CHOEP-14 or to dose-escalated sequential high-dose therapy plus rituximab (R-MegaCHOEP) with serial autologous HCT [44]. Severe (grade 3/4) hematologic and non-hematologic toxicities were more common following R-MegaCHOEP. At a median follow-up of 42 months, when compared with R-MegaCHOEP, R-CHOEP-14 resulted in similar estimated three-year EFS (70 versus 61 percent) and OS (74 versus 70 percent).

We do not advocate the use of autologous HCT in first remission for most patients with DLBCL, including those with high-intermediate-risk or high-risk IPI scores. While a post hoc analysis of the S9704 intergroup trial suggests that autologous HCT is an option for selected patients with high-risk (not high-intermediate-risk) IPI, we believe that, given the uncertainty of this benefit, it does not outweigh the known morbidities associated with HCT for most patients. Autologous HCT is still considered for the treatment of relapsed disease and its use in this setting is discussed in detail separately. (See "Diffuse large B cell lymphoma (DLBCL): Suspected first relapse or refractory disease in medically-fit patients", section on 'Autologous HCT'.)

Activated B cell type DLBCL — Together, ABC DLBCL and non-GCB DLBCL account for approximately 60 percent of de novo DLBCL, although the frequency increases with increasing age [27]. For patients with advanced stage ABC type DLBCL, we encourage enrollment on a clinical trial. Additional information and instructions for referring a patient to an appropriate research center can be obtained from the United States National Institutes of Health (www.clinicaltrials.gov). The more intensive R-ACVBP regimen (rituximab, doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone, followed by consolidation with methotrexate and leucovorin) is an acceptable alternative for younger patients (<60 years of age) in practices that have access to vindesine.

Among patients with advanced stage ABC DLBCL, standard therapy with R-CHOP-21 results in estimated rates of PFS and OS at five years of 45 and 50 percent, respectively [22]. Subgroup analyses of prospective trials suggest that alternative regimens result in superior outcomes. Of these, only R-ACVBP has demonstrated superiority to R-CHOP-21 in a randomized trial. However, R-ACVBP incorporates vindesine, an agent that is not available in the United States. Support for the other regimens (R-CHOP plus bortezomib or lenalidomide) comes from analyses that suggest the addition of these novel agents abrogates the inferior outcomes seen in the non-GCB DLBCL subtype.

●R-ACVBP – A more intensive rituximab-based regimen (R-ACVBP) has demonstrated superior DFS and OS to R-CHOP in a single prospective randomized trial when used for a highly select group of young patients with DLBCL [45]. This trial randomly assigned 379 younger patients (<60 years of age) with newly diagnosed DLBCL and an age-adjusted IPI of one to either four cycles of R-ACVBP (rituximab, doxorubicin, cyclophosphamide, vindesine, bleomycin, and prednisone, followed by consolidation with methotrexate and leucovorin) or eight cycles of R-CHOP-21. When compared with R-CHOP-21, R-ACVBP resulted in a similar CR rate (83 versus 80 percent), and superior three-year EFS (81 versus 67 percent) and OS (92 versus 84 percent) rates. R-ACVBP was associated with more hematologic and non-hematologic toxicity. Further analysis of this study suggested that the superiority of R-ACVBP over R-CHOP was limited to tumors with immunohistochemical features suggesting a non-germinal center B cell-like phenotype (negative for both CD10 and BCL6 or CD10–/BCL6+/MUM1+) [46].

●R-CHOP plus lenalidomide (R2-CHOP) – An open-label, single arm, multicenter trial (REAL07) evaluated the use of R-CHOP plus lenalidomide (R2-CHOP) in 49 older adults (age 60 to 80 years) with previously untreated advanced stage DLBCL or grade 3b follicular lymphoma [47]. Treatment included six cycles of oral lenalidomide (15 mg on days 1 to 14) in combination with standard R-CHOP-21. The overall response rate and CR rate were 92 and 86 percent, respectively. Severe (grade 3/4) toxicity included neutropenia (31 percent), leukopenia (28 percent), and thrombocytopenia (13 percent). After a median follow-up of 28 months, the estimated PFS and OS rates at two years were 80 and 92 percent. Results were similar among those with GCB and ABC subtypes.

In another phase II trial of R2-CHOP, 64 patients with previously untreated advanced stage DLBCL were treated with oral lenalidomide (25 mg on days 1 through 10) plus standard-dose R-CHOP-21 administered for six cycles [48]. Of the 60 patients available for response assessment, the overall response rate was 98 percent (80 percent complete). Estimated EFS and OS rates at 24 months were 59 and 78 percent, respectively. EFS rates were similar for those with GCB and ABC subtypes. Hematologic toxicity was universal. Severe (grade 3/4) nonhematologic toxicities were seen in 25 percent of patients. One patient died from sepsis following gut perforation at a site of DLBCL involvement after the first cycle of R2-CHOP. Three patients developed a second malignancy (acute myeloid leukemia, glioblastoma, and metastatic colon cancer).

●R-CHOP plus ibrutinib – A multicenter trial reported no difference after median 35 months in EFS or OS among 838 patients with non-GCB DLBCL (76 percent with ABC subtype) who were randomly assigned to R-CHOP plus ibrutinib versus R-CHOP alone [49]. In patients <60 years, ibrutinib plus R-CHOP was associated with improved EFS, PFS, and OS with manageable toxicity, but in patients ≥60 years, the combination was associated with increased toxicity, compromised administration of R-CHOP, and inferior outcomes.

●R-CHOP plus bortezomib – R-CHOP plus bortezomib is an investigational combination. A single arm phase II trial evaluated the use of R-CHOP plus bortezomib in 40 patients with previously untreated DLBCL [50]. A partial or better response was seen in 88 percent with 75 percent achieving a CR or CR unconfirmed. At a median follow-up of 51 months, two-year rates of PFS and OS were 64 and 70 percent, respectively. The addition of bortezomib was associated with higher than usual rates of peripheral neuropathy. Of interest, there was no survival difference observed between patients with GCB or ABC gene expression profiles (GEPs), suggesting that the inferior outcomes normally seen in patients with ABC DLBCL may be abrogated with the addition of bortezomib. In contrast, there was no benefit seen with the substitution of vincristine in R-CHOP with bortezomib (VR-CAP) in a randomized trial of 164 patients with non-GCB DLBCL [51]. The VR-CAP arm required more significant dose reductions resulting in an inferior dose-intensity in this population. Further studies are needed to evaluate the use of bortezomib in this population before it can be incorporated into clinical care.  

Double hit or double expressor DLBCL — Double hit DLBCL (ie, lymphomas with rearrangements of MYC and BCL2 and/or BCL6) is categorized in the 2016 World Health Organization classification as high-grade B cell lymphoma, with MYC and BCL2 and/or BCL6 rearrangements [52]. Other cases of de novo DLBCL are described as double expressor DLBCL (ie, expression of both MYC and BCL2 by immunohistochemistry, but without rearrangements of the MYC and BCL2 genes). Evidence regarding the poor prognosis is clearer for patients with double hit DLBCL than for the larger population with double expressor DLBCL, for whom estimated rates of PFS and OS for all stages of disease following standard R-CHOP-21 are approximately 40 to 50 and 40 to 55 percent, respectively [22,25]. (See "Prognosis of diffuse large B cell lymphoma", section on 'Identifying molecular subtype'.)

For patients with advanced stage double hit or double expressor DLBCL, we encourage enrollment in a clinical trial. Additional information and instructions for referring a patient to an appropriate research center can be obtained from the United States National Institutes of Health (www.clinicaltrials.gov).

For patients treated outside of the context of a clinical trial, we suggest:

●Double expressor DLBCL – R-CHOP is the most appropriate treatment, given that there are no trials in this setting that have directly compared R-CHOP with more intensive regimens.

●Double hit DLBCL – More intensive regimens (see below) provide superior clinical outcomes in double hit lymphoma when directly compared with R-CHOP. We generally treat with six to eight cycles of dose-adjusted (da)-EPOCH-R with a goal of administering two cycles beyond attainment of a complete remission.

Acceptable intensive chemotherapy regimens for treatment of double hit DLBCL include:

•da-EPOCH-R (etoposide, doxorubicin, vincristine, cyclophosphamide, and prednisone plus rituximab) (table 3)

•R-HyperCVAD/MA (rituximab plus hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone alternating with methotrexate and cytarabine) (table 4)

•R-CODOX-M/IVAC ("Magrath regimen"; rituximab plus cyclophosphamide, vincristine, doxorubicin, and high-dose methotrexate alternating with ifosfamide, etoposide, and cytarabine) (table 5)

Evidence in support of more intensive regimens for double hit lymphoma includes the following:

●A 2015 systematic review and meta-analysis included data from 11 retrospective studies, including a total of 394 patients with double hit lymphoma [53]. When compared with R-CHOP, treatment with da-EPOCH-R reduced the risk of progression (median 22 versus 12 months, hazard ratio [HR] 0.66; 95% CI 0.44-0.96), but did not translate into improved OS (median 31 versus 21 months; HR 0.77, 95% CI 0.51-1.13).

●A multicenter study of 311 patients reported that compared with R-CHOP, da-EPOCH-R achieved superior PFS, but not improved OS [54].

●A multicenter study that included 159 patients reported that compared with R-CHOP, more intensive front-line therapy (eg, da-EPOCH-R, R-HyperCVAD/MA, R-CODOX-M/IVAC) was associated with superior three-year relapse-free survival (RFS; 88 versus 56 percent, respectively) [55]. Three-year RFS and OS for the entire cohort were 80 and 87 percent, respectively.

A benefit for autologous HCT in first complete remission (CR1) has not been demonstrated. Examples of studies that have examined this question include the following:

●In a large multicenter study, autologous HCT was not associated with improved survival among patients attaining a complete response [54].

●In another multicenter study, RFS and OS did not differ between those who did or did not undergo autologous HCT in CR1 [55].

While a benefit for autologous HCT in this setting is not proven, younger patients in CR1 should be offered the opportunity to discuss the risks and benefits of autologous HCT with a transplant physician. (See "Determining eligibility for autologous hematopoietic cell transplantation".)  

SPECIAL SCENARIOS — The general treatment principles for limited and advanced stage DLBCL apply to most disease presentations and patient populations. However, involvement of certain organs, particularly when occurring as primary sites of disease, has unique therapeutic implications. In addition, older adults and patients with cardiac disease or human immunodeficiency virus (HIV) infection require special consideration.

Older adults — Older adults (ie, >60 years) with DLBCL generally have a worse prognosis compared to younger patients due, in part, to more comorbid conditions, lower treatment tolerance, and administration of inadequate therapy. A comprehensive geriatric assessment may be useful in assessing comorbid conditions and functional status in the older patient, thereby permitting formulation of an appropriate, individualized treatment plan [56-59]. Special considerations for the use of chemotherapy in the elderly population are discussed separately. (See "Comprehensive geriatric assessment for patients with cancer" and "Systemic chemotherapy for cancer in older adults".)

We suggest treating most adults >60 years of age with full-dose therapy rather than lower doses; adjustments can be made to future cycles according to tolerance [57,60-62]. With a potentially curable disease such as DLBCL, it is important to treat all patients (including older individuals) with planned therapy whenever possible, because the amount of chemotherapy administered during initial treatment affects overall survival (OS). (See 'Dose intensity' above.)

Retrospective analysis of more than 600 patients with DLBCL ≥60 years of age who were treated with full dose R-CHOP every 14 days or every 21 days reported five-year OS of 69 percent (95% CI 65 to 73 percent) [63]. With either treatment schedule, older patients received ≥98 percent median planned doses for all agents. Toxicity profiles were comparable in both treatment arms, with any grade ≥3 adverse events seen in approximately two-thirds of patients. There was a higher incidence of neutropenia (61 versus 36 percent, respectively) with R-CHOP-21 compared with R-CHOP-14 (presumably due to reduced use of G-CSF with R-CHOP-21), but less thrombocytopenia (7 versus 12 percent, respectively). The incidence of deaths during chemotherapy was less than 2 percent with both treatment regimens.

This approach may be confounded by comorbidities that independently affect survival and can limit total doses of chemotherapy. Treatment with a reduced-dose CHOP regimen (eg, R-mini-CHOP; rituximab 375 mg/m², cyclophosphamide 400 mg/m², doxorubicin 25 mg/m², and vincristine 1 mg on day 1 of each cycle, and 40 mg/m² prednisone on days 1 to 5) may be considered in older adults who are unable to tolerate standard doses of R-CHOP-21 [64]. (See 'Germinal center B cell type DLBCL' above.)

An analysis of the Surveillance, Epidemiology, and End Results (SEER) database identified 1156 patients >80 years old who had DLBCL [65]. When compared with other patients with DLBCL, these older adults were less likely to receive R-CHOP and were more likely to receive R-CVP or to be observed. On multivariable analysis, R-CHOP was the only treatment regimen associated with improved OS (hazard ratio [HR] 0.45; 95% CI 0.33-0.63) and lymphoma-related survival (HR 0.58; 95% CI 0.38-0.88).

Older adults have greater variability in drug pharmacokinetics, which may also differ by sex. In particular, when compared with older men and younger individuals of both sexes, older women have reduced rituximab clearance, resulting in higher serum levels and prolonged exposure times. A post hoc analysis of patients enrolled on six prospective clinical trials of rituximab reported that the incremental benefit attributed to the addition of rituximab was greatest among older females [66]. This inflated benefit was thought to be due to prolonged exposure times and suggested that the standard dosing may be suboptimal in other populations.

A phase II trial (SMARTE-R-CHOP-14) evaluated the addition of eight extra rituximab doses to six cycles of R-CHOP-14 in 189 older adults (age 61 to 80 years) with DLBCL [67]. A complete response (CR) or unconfirmed CR was achieved in 85 percent. Three-year event-free survival (EFS) and OS rates were 71 and 84 percent, respectively. For a population as a whole, the clinical outcomes were similar to those seen in historical controls treated with six cycles of R-CHOP-14 (patients from RICOVER-60 [7]). For the subpopulation of older adults with a high International Prognostic Index score (IPI 3 to 5) (table 6), the additional doses of rituximab were associated with improved three-year EFS (67 versus 54 percent) and OS (80 versus 67 percent).

At this time, rituximab dosing does not differ among populations. Prospective trials are necessary to determine the optimal dosing of rituximab in different patient populations.

There is a paucity of data regarding the treatment of the very elderly (ie, >80 years old) adult with DLBCL, but a subset of these patients may be cured with less intensive chemotherapy plus rituximab [64,68-71]. It is possible that the addition of rituximab in this setting offsets the potential negative effect of chemotherapy dose reduction, thereby providing acceptable efficacy and decreased toxicity. A multicenter phase II study evaluated six cycles of R-mini-CHOP in 150 patients age 80 years or older with DLBCL (75 percent stage III or IV) [64]. Hematologic toxicity was the most common severe (grade 3/4) side effect with neutropenia and thrombocytopenia rates of 39 and 7 percent, respectively. Eleven patients (7 percent) experienced at least one episode of febrile neutropenia. There were 12 deaths related to toxicity. The overall response rate was 74 percent (63 percent CR or unconfirmed CR). At a median follow-up of 20 months, the rates of OS and progression-free survival (PFS) at two years were 59 and 47 percent, respectively.

In certain circumstances, a patient may be judged to be too frail to withstand even reduced dose combination chemotherapy (eg, R-mini-CHOP). In such a setting, palliation of symptoms may be achieved by sequential use of single chemotherapeutic agents with or without steroids.

Patients with cardiac disease — Although anthracycline-based therapy, such as CHOP, is recommended for patients with DLBCL, patients with underlying cardiac disease may not be able to tolerate the use of an anthracycline since this agent is toxic to cardiac cells [72]. Doxorubicin or other anthracyclines should not be administered to patients with a baseline ejection fraction below 30 percent. Patients with an ejection fraction greater than 30 percent may require monitoring. Contrary to common belief, mitoxantrone is not a safe alternative in patients with cardiac disease and produces inferior results [17]. However, the use of pegylated liposomal doxorubicin may permit substantially higher cumulative doses with efficacy equivalent to the free anthracyclines but a lower incidence of heart failure [73]. The use of anthracyclines in patients with cardiac disease is discussed in detail separately. (See "Clinical manifestations, monitoring, and diagnosis of anthracycline-induced cardiotoxicity" and "Prevention and management of anthracycline cardiotoxicity".)

When an anthracycline is not an option, non-anthracycline-containing combination chemotherapy programs must be used. When choosing among such regimens, one must keep in mind not only the potential cardiac toxicity but also the intravenous fluid requirements of a regimen. As an example, cisplatin-containing regimens, although active in DLBCL, may require the administration of too much fluid to be easily used for patients with cardiac disease.

There is a paucity of data to guide the choice of therapy. There are no randomized trials of chemotherapy for DLBCL that included patients with underlying cardiac disease. The following are the largest prospective trials in this population:

●R-GCVP – A nonrandomized phase II trial investigated the use of rituximab, gemcitabine, cyclophosphamide, vincristine, and prednisolone (R-GCVP) in 62 adults (median age 77 years) with advanced stage DLBCL considered unfit for anthracycline-containing chemotherapy because of cardiac comorbidity [74]. Ischemic heart disease was present in 60 percent, and 44 percent had a left ventricular ejection fraction ≤50 percent. The majority (71 percent) received greater than three cycles, and approximately half (52 percent) were able to complete six cycles. Severe (grade ≥3) toxicity was seen in 56 percent with the most common being neutropenia (29 percent), thrombocytopenia (19 percent), infection (17 percent), fatigue (14 percent), and cardiac toxicity (10 percent). There were three cardiac-related deaths. The overall response rate was 61 percent (39 percent CR or unconfirmed CR). Median PFS was 17 months and 56 percent were alive at two years.

●CEPP – A nonrandomized phase II trial investigated the use of cyclophosphamide, etoposide, prednisone, procarbazine (CEPP) in 83 patients (mostly <60 years old) with aggressive non-Hodgkin lymphoma, 14 of whom had contraindications to doxorubicin [75]. The therapy was well tolerated with no cardiac toxicity. The most common side effects were neutropenic fever (12 percent) and severe nausea and vomiting (7 percent). Of the previously untreated patients with DLBCL, complete and partial responses were seen in nine (64 percent) and two (14 percent). Of importance, this study was primarily conducted in younger adults with cardiac disease. Toxicities associated with procarbazine increase with age such that it may be difficult to administer this regimen to patients over the age of 60. There have been no published reports of this regimen in combination with rituximab. However, we advocate the use of rituximab with this regimen since rituximab has dramatically improved results with other regimens and there are no common toxicities between rituximab and CEPP.

These regimens have not been compared directly in a randomized trial. In addition, response rates reported in these trials cannot be accurately compared due to changes over time in the response assessment and response definitions. As such, a choice among chemotherapy regimens in this setting is primarily made based upon the clinician's experience with the regimen and expected toxicity. For patients who are unable to receive an anthracycline, any of these regimens would be acceptable.

AIDS-related lymphoma — Among patients with HIV, the diagnosis of non-Hodgkin lymphoma is an AIDS-defining malignancy. A significant number of the lymphomas encountered in HIV-infected patients are clinically aggressive (eg, DLBCL) or highly aggressive (eg, immunoblastic or small, non-cleaved cell/Burkitt's-like lymphoma). Treatment of such patients is complicated by their immunocompromised state and also requires specific treatment for their HIV. As such, treatment of these patients is discussed in detail separately. (See "HIV-related lymphomas: Treatment of systemic lymphoma".)

Bulky disease — The role for radiation in bulky sites of advanced disease is less certain. It is unclear if patients with advanced disease with bulky lesions are more likely to relapse in the area of previous bulky disease or elsewhere. Local relapse could theoretically be decreased by radiation, but such treatment is unlikely to decrease systemic relapse [76-78]. In the absence of clear benefit in the rituximab era, we do not advocate radiation therapy after chemotherapy in patients with bulky components to otherwise advanced disease.

Retrospective and observational studies have had mixed results [76,77,79,80]. As examples:

●A retrospective series that included 24 patients with bulky stage III/IV disease reported that patients with advanced disease who had received CHOP without radiation developed relapse in previously bulky sites only 10 percent of the time [76]. The other 90 percent of relapses occurred at distant sites.

●Another single institution retrospective analysis of 469 patients with newly diagnosed DLBCL, the majority of whom were treated with initial R-CHOP chemotherapy with or without radiation, included 207 patients with bulky disease >5 cm [77]. Of the 279 patients with advanced stage disease, 39 were given radiation therapy and 23 of these 39 patients had bulky disease. The small number of patients with advanced stage DLBCL who received radiation therapy had superior rates of five-year OS (89 versus 66 percent) and PFS (76 versus 55 percent) when compared with those who did not receive radiation. On matched-pair analysis, bulky disease status did not affect the outcome in relation to the role of radiation therapy.

●A post hoc combined analysis of the RICOVER-60 trial and an expansion cohort evaluated the impact of radiation therapy on the outcomes of 164 older adults (61 to 80 years) with initial bulky disease (≥7.5 cm) treated with six cycles of R-CHOP-14 [81]. Radiation therapy was not randomized. For patients enrolled on RICOVER-60, radiation therapy (36 Gy) was specified for all patients with bulky disease or extralymphatic disease regardless of the result of chemotherapy. The expansion cohort was enrolled to receive the same chemotherapy without radiation. This group was not matched for clinical risk factors. When analyzed by intention-to-treat, the omission of radiation decreased EFS (hazard ratio 2.1) and was associated with a trend toward inferior PFS and OS, which reached statistical significance in a per protocol analysis.

There have been no prospective, randomized trials of adjuvant radiation therapy for bulky advanced stage disease in the rituximab era. Some clinicians advocate radiation to bulky sites after a complete remission is obtained with chemotherapy [82].

DLBCL-Burkitt — DLBCL, not otherwise specified (NOS) is a category in the 2016 World Health Organization (WHO) classification [52] that includes lymphomas with features intermediate between DLBCL and Burkitt lymphoma, but do not truly fit into either category. The colloquial term "gray zone lymphoma" has been used in the past to include this entity. This category of DLBCL accounts for approximately 5 percent of cases previously described as DLBCL [83].

Many cases with this immunophenotype are "double hit" lymphomas, and, as such, the 2016 WHO classification describes them as high-grade B cell lymphoma, with MYC and BCL2 and/or BCL6 rearrangements (table 7) [52]. (See 'Double hit or double expressor DLBCL' above.)

Outcomes with R-CHOP-21 are disappointing with a median OS of only 2.5 to 18 months [84-87]. In addition, the vast majority of patients with this entity are older adults (median age 70 years) and are not candidates for highly intensive therapy. There is a paucity of data regarding other treatment options for such patients, and they should be encouraged to participate in clinical trials. Outside of a clinical trial, we generally treat with da-EPOCH-R (dose-adjusted cyclophosphamide, doxorubicin, etoposide, vincristine, prednisone plus rituximab) given its activity in both Burkitt lymphoma and DLBCL and its reasonable tolerability (table 3) [54,88,89].

While some experts suggest the use of aggressive combination chemotherapy regimens commonly used for the treatment of Burkitt lymphoma, such as CODOX-M/IVAC (cyclophosphamide, vincristine, doxorubicin, high-dose methotrexate with ifosfamide, cytarabine, etoposide, and intrathecal methotrexate) (table 5), toxicity with this regimen is severe and virtually all patients require a prolonged inpatient hospital stay and blood product support. As such, we generally reserve the use of these intensive regimens for younger patients with a good performance status.

DLBCL-Hodgkin — Occasional cases of lymphoma have clinical and pathologic features intermediate between DLBCL and classic Hodgkin lymphoma; such cases are described in the 2016 WHO classification as B cell lymphoma, unclassifiable, with features intermediate between DLBCL and classic Hodgkin lymphoma (table 7). The colloquial term "gray zone lymphoma" has been used in the past to include this entity. The prognosis and optimal treatment of such cases are uncertain. In general, we use a similar approach to that used for primary mediastinal large B cell lymphoma, which is described in more detail separately. (See "Primary mediastinal large B cell lymphoma".)

PATIENT FOLLOW-UP — After completion of the initially planned treatment of DLBCL, patients should be evaluated to determine the disease response to treatment and should be followed longitudinally for relapse.

Response evaluation — One month following the completion of planned therapy (or sooner if the outcome is unfavorable), the response to treatment should be documented by history, physical examination, and laboratory studies (complete blood count, lactate dehydrogenase [LDH], and biochemical profile). The post-treatment imaging study of choice is the positron emission tomography with computed tomography (PET/CT) scan, which provides information on the size and activity of residual masses and allows for the distinction between active disease and fibrosis. PET/CT should be obtained six to eight weeks after completion of chemotherapy and 12 weeks after the completion of radiation therapy [90]. PET/CT imaging obtained earlier than this is likely to demonstrate increased uptake due to an inflammatory reaction to treatment.

Using information gathered from the history, physical, and PET/CT scan, disease response is determined (table 8). A discussion of the accuracy of PET/CT in the follow-up of patients with lymphoma is presented separately. (See "Pretreatment evaluation, staging, and response assessment of non-Hodgkin lymphoma", section on 'Evaluating response to treatment'.)

Patients who fail to obtain a complete response are treated as refractory disease (after biopsy-proven persistent disease). Treatment of refractory disease is presented in detail separately. (See "Diffuse large B cell lymphoma (DLBCL): Suspected first relapse or refractory disease in medically-fit patients".)

We do not advocate the use of PET scans obtained prior to the completion of therapy (ie, interim PET scans) unless there is clinical deterioration necessitating evaluation. This is principally because there are limited data regarding the predictive value of such scans. While some studies have suggested that positive interim PET scans are associated with inferior outcomes [91], there appears to be substantial inter-reader variability, even among nuclear medicine experts. As an example, a prospective trial that evaluated the interpretation of 38 interim PET scans reported complete agreement by Eastern Cooperative Oncology Group (ECOG) and London criteria in 68 and 71 percent, respectively [92].

Surveillance for relapse — Following the completion of therapy, restaging, and documentation of complete remission, patients are seen at periodic intervals to monitor for treatment complications and assess for possible relapse. The frequency and extent of these visits depends upon the comfort of both the patient and physician. There have been no prospective, randomized trials comparing various schedules of follow-up. Our approach is based upon the following general understandings:

●The majority of relapses occur during the first two years after completion of treatment [93,94].

●Relapses are usually symptomatic and are rarely identified solely on the basis of routine imaging [95,96].

●If a relapse is picked up a few weeks earlier because of more intense monitoring, it is unlikely to improve outcome.

●When planning the post-treatment surveillance strategy, care should be taken to limit the number of CT scans, particularly in younger individuals, given concerns about radiation exposure and the risk for second malignancies [97-99]. (See "Radiation-related risks of imaging".)

Our approach to patient surveillance is to schedule patient visits every three months during the first year, every three to six months during the second year, and every six months starting two years after complete response. At these visits, we perform a history and physical examination, complete blood count, chemistries, and LDH.

There is no role for routine PET or PET/CT imaging in the longitudinal follow-up of asymptomatic patients after response assessment. There are limited data to support performing CT scans at a given interval, and the decision regarding frequency of CT imaging should be individualized to the patient. For most patients, we do not routinely perform surveillance imaging. In a retrospective study, 112 of 552 (20 percent) patients with DLBCL relapsed after achieving an initial remission with induction therapy [100]. The majority of relapses (64 percent) were identified prior to a scheduled follow-up visit. Only nine relapses (1.6 percent) were identified by surveillance imaging before the development of clinical manifestations. There was no difference in survival after relapse among those detected by imaging versus clinical manifestations. Another population-based study found similar survival rates among patients managed with similar follow-up strategies that either included (Danish registry) or did not include (Swedish registry) routine CT imaging [101].  

Noninvasive monitoring with immunoglobulin high-throughput sequencing may allow for a highly specific, radiation-free monitoring technique in the future [102-104]. This method searches for clone-specific DNA sequences in the peripheral blood.

Relapsed disease can be suggested by changes on imaging studies but can only be confirmed by biopsy. As such, a biopsy should always be obtained to document relapsed disease before proceeding to salvage therapy. The treatment of relapsed DLBCL is presented separately. (See "Diffuse large B cell lymphoma (DLBCL): Suspected first relapse or refractory disease in medically-fit patients".)

Prognosis — DLBCL is a heterogeneous group of clinically aggressive non-Hodgkin lymphoma. Without treatment, survival is measured in months. A percentage of patients can be cured following immunochemotherapy with or without radiation therapy. In addition to stage of disease, survival with immunochemotherapy varies with the presence or absence of certain clinical features and molecular/genetic features.

The original and revised International Prognostic Index (IPI) scores provide information on patient outcomes based upon the patient's age, performance status, stage, number of extranodal sites, and LDH level (table 6 and table 9). Estimated survival rates at four years range from >90 percent for those with low scores to <60 percent for those with high scores.

The genetic and molecular profile of the tumor can also predict patient outcome following standard therapy. The following groups have distinct prognoses:

●Double hit DLBCL is defined as those having MYC translocation plus gene rearrangement of BCL2, BCL6, or both. The approximately 10 percent of cases with double hit biology often demonstrate early progression and refractory disease following R-CHOP. Initial reports suggest a similarly poor prognosis among those with double protein DLBCL, as defined as cases in which immunohistochemistry identifies co-expression of MYC and BCL2, but gene rearrangements are not present or not evaluated.

●Germinal center B cell (GCB) type DLBCL has a gene expression profile (GEP) that resembles a normal germinal center B cell and has excellent rates of five-year survival following standard therapy with R-CHOP.

●Activated B cell (ABC) type DLBCL has a GEP that resembles an activated B cell. When compared with GCB tumors, ABC tumors are associated with inferior survival.

Further information regarding the prognosis of DLBCL is presented separately. (See "Prognosis of diffuse large B cell lymphoma".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Management of diffuse large B cell lymphoma".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topics (see "Patient education: Diffuse large B cell lymphoma (The Basics)" and "Patient education: What are clinical trials? (The Basics)" and "Patient education: Neutropenia and fever in people being treated for cancer (The Basics)" and "Patient education: Nausea and vomiting with cancer treatment (The Basics)" and "Patient education: Hair loss from cancer treatment (The Basics)" and "Patient education: When your cancer treatment makes you tired (The Basics)")

●Beyond the Basics topics (see "Patient education: Diffuse large B cell lymphoma in adults (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Diffuse large B cell lymphoma (DLBCL) is the most common histologic subtype of non-Hodgkin lymphoma (NHL). It is an aggressive NHL in which survival without treatment is measured in months. Advanced stage refers to disease that cannot be contained within one irradiation field, and accounts for 60 to 70 percent of patients with DLBCL. (See 'Introduction' above.)

●Pretreatment evaluation determines the extent of the disease and provides information about the individual's comorbidities that are likely to affect treatment options. In addition to a history and physical examination, it is our practice to perform laboratory studies, unilateral bone marrow biopsy, and imaging in all patients. Certain subsets of patients will require further testing of their cerebrospinal fluid and cardiac function. Fertility counseling should be offered to patients in childbearing years. (See 'Pretreatment evaluation' above.)

●There is broad consensus that initial treatment of all patients with DLBCL should include rituximab (or another anti-CD20 monoclonal antibody). For all patients with advanced stage DLBCL, we recommend treatment with anti-CD20 immunotherapy plus chemotherapy rather than chemotherapy alone (Grade 1A), based on randomized clinical trials that demonstrated superior survival with little additional toxicity. (See 'Targeting CD20' above.)

●A molecular risk assessment should be performed in all cases to help determine prognosis and direct therapy. This includes both an evaluation of MYC, BCL2, and BCL6 status (by immunohistochemistry or fluorescence in situ hybridization [FISH]) and an evaluation of cell of origin (by gene expression profiling [GEP], immunohistochemistry-based algorithms (figure 1), or LymphCx platform). Using this information, an individual case may be subclassified as one of the following (see 'Identify disease subtype' above):

•Germinal center B cell (GCB) DLBCL – Cases with GCB DLBCL are those identified by GEP, immunohistochemistry algorithms, or LymphCx without MYC and BCL2 gene rearrangements. These patients have a relatively good prognosis following standard therapy with R-CHOP. For most patients with advanced stage GCB DLBCL, we recommend treatment with an anthracycline-based combination chemotherapy regimen plus rituximab (Grade 1A). Specifically, we use six cycles of R-CHOP-21 (table 2). There is no role for the routine use of maintenance rituximab or high-dose chemotherapy with autologous hematopoietic cell rescue in first complete remission. (See 'Germinal center B cell type DLBCL' above.)

•Activated B cell (ABC) DLBCL or non-GCB DLBCL – Cases with non-GCB DLBCL identified by GEP, immunohistochemistry algorithms, or LymphCx, without double hit DLBCL, have high relapse rates and a poor prognosis following treatment with R-CHOP. For patients with advanced stage ABC type DLBCL, we encourage enrollment on a clinical trial evaluating the incorporation of novel agents (eg, R-CHOP plus lenalidomide; R-CHOP plus ibrutinib; R-CHOP plus bortezomib). The more intensive R-ACVBP regimen is an acceptable alternative in practices that have access to vindesine. (See 'Activated B cell type DLBCL' above.)

•Double hit DLBCL – Cases with MYC translocation plus gene rearrangement of BCL2, BCL6 (or both) have a poor prognosis with standard therapy. These patients should be encouraged to enroll on a clinical trial. For those treated off study, we suggest six to eight cycles of dose-adjusted EPOCH-R (etoposide, doxorubicin, vincristine, cyclophosphamide, and prednisone plus rituximab) (table 3) with a goal of administering two cycles beyond attainment of a complete remission (Grade 2C). (See 'Double hit or double expressor DLBCL' above.)

•Double expressor DLBCL – There are limited data regarding the treatment of the larger population of patients with double expressor DLBCL (immunohistochemistry identifies co-expression of MYC and BCL2, but gene rearrangements are not present or were not evaluated). Such patients should be encouraged to enroll on a clinical trial or treated with R-CHOP off study. (See 'Double hit or double expressor DLBCL' above.)

●The general treatment principles for limited and advanced stage DLBCL apply to most disease presentations and patient populations. However, there are unique therapeutic implications regarding involvement of certain organs, particularly when occurring as primary sites of disease. In addition, older adults and patients with cardiac disease or human immunodeficiency virus (HIV) infection require special consideration. (See 'Special scenarios' above.)

For most older adults (age >60 years) with DLBCL we suggest full-dose therapy rather than lower doses (Grade 2C). Older adults who are unable to tolerate standard doses of R-CHOP-21 may be considered for a reduced-dose CHOP regimen (ie, R-mini-CHOP). (See 'Older adults' above.)

●Patients with disease affecting certain organs (eg, testicular, epidural, or sinus involvement) are at increased risk of developing central nervous system (CNS) disease. This is discussed in more detail separately. (See "Clinical presentation and diagnosis of secondary central nervous system lymphoma", section on 'Incidence and risk factors'.)

●Disease response is determined using information from positron emission tomography with computed tomography (PET/CT, which should be obtained six to eight weeks after chemotherapy or at least 12 weeks after the completion of radiation therapy), along with post-treatment history, physical exam, and laboratory studies (table 8). (See 'Response evaluation' above.)

Patients are seen at periodic intervals to monitor for treatment complications and assess for possible relapse. The frequency and extent of these visits depends upon the comfort of both the patient and physician. (See 'Surveillance for relapse' above.)