Diffuse large B-cell lymphoma (DLBCL) is an aggressive lymphoma arising from mature B lymphocytes, the cells of the immune system that normally participate in antibody-mediated defense. Its name describes its microscopic appearance: large malignant lymphoid cells grow in a diffuse pattern, replacing the normal organization of affected tissue. DLBCL is the most common subtype of non-Hodgkin lymphoma in adults in many populations. It usually develops rapidly, but unlike many advanced solid cancers, it remains potentially curable even when several organs are involved.
DLBCL is not a single biologically uniform disease. Tumors with similar appearances can differ in their developmental origins, genetic alterations, interactions with surrounding cells, and responses to treatment. Diagnosis therefore integrates tissue morphology, immunologic markers, and selected molecular tests. Treatment has evolved from combination chemotherapy to antibody-based regimens and, for recurrent or resistant disease, cellular therapies and other targeted agents.
Definition, classification, and historical development
DLBCL belongs to the mature B-cell neoplasms. Normal B cells undergo several stages of development and activation, including changes that improve their ability to recognize foreign antigens. A malignant clone can acquire alterations during these processes and expand in lymph nodes or extranodal tissues. In DLBCL, the malignant cells are generally larger than ordinary resting lymphocytes, with prominent nuclei, visible nucleoli, and a relatively high rate of division. “Diffuse” refers to the absence of the organized follicular growth pattern characteristic of some other B-cell lymphomas; it does not mean that the disease must be widespread throughout the body.
The designation DLBCL, not otherwise specified, abbreviated DLBCL, NOS, applies when a large B-cell lymphoma does not meet the defining criteria for a more specific entity. Modern classifications distinguish several related diseases by anatomic setting, viral association, molecular features, or characteristic clinical behavior. Examples include primary mediastinal large B-cell lymphoma, intravascular large B-cell lymphoma, primary cutaneous DLBCL, leg type, and Epstein–Barr virus-positive DLBCL. Primary central nervous system lymphoma is also distinguished from ordinary systemic DLBCL because of its site, biology, and treatment requirements. These entities cannot be treated as interchangeable merely because their cells look similar.
Some diagnostic boundaries have changed as genetic evidence has accumulated. The fifth edition of the World Health Organization classification and the 2022 International Consensus Classification recognize a distinct category for tumors with rearrangements involving both MYC and BCL2, although their terminology and some boundaries differ. Tumors with MYC and BCL6 rearrangements alone are no longer automatically grouped with MYC/BCL2-rearranged disease in the WHO system. Consequently, older studies using the broad term “double-hit lymphoma” may combine diseases that current classifications separate.
DLBCL may arise as a new malignancy or develop through transformation of an indolent lymphoma, particularly follicular lymphoma. Transformation of chronic lymphocytic leukemia into an aggressive lymphoma, most often DLBCL, is called Richter transformation. These settings influence biological interpretation and clinical expectations. Historically, the development of immunophenotyping helped replace older morphology-dominated classifications, while the 1994 Revised European-American Lymphoma classification and subsequent WHO classifications established increasingly reproducible disease categories. The addition of rituximab to chemotherapy around the turn of the twenty-first century marked a separate therapeutic transition, demonstrating that targeting a B-cell surface antigen could substantially improve survival.
Epidemiology and associated conditions
DLBCL accounts for approximately 25–35% of adult non-Hodgkin lymphomas in many Western populations, although its proportion varies with geography, age structure, and the frequency of other lymphoma subtypes. Its incidence rises strongly with age, and the median age at diagnosis in many large series is in the mid-to-late sixties. It nevertheless occurs in young adults and, less commonly, children. Men are affected somewhat more often than women. Comparisons between countries require care because cancer-registration coverage, pathology resources, and classification practices influence reported rates.
Most patients have no single identifiable cause. The disease generally results from acquired genetic changes rather than inheritance of a lymphoma-causing mutation. Familial aggregation of lymphoid cancers has been observed, but most DLBCL cases occur without a strong family history. Age-related changes in immune function and the accumulation of cellular genetic damage provide part of the background in which the disease develops. Population associations do not establish that any particular exposure caused an individual tumor.
Conditions that impair immune surveillance increase the risk of aggressive B-cell lymphomas. HIV infection, inherited immune deficiencies, and prolonged immunosuppression following organ transplantation are important examples. DLBCL is a common histologic pattern within post-transplant lymphoproliferative disorders, although those disorders form a broader clinicopathologic category rather than a simple synonym for ordinary DLBCL. Effective antiretroviral treatment has changed the epidemiology and management of HIV-associated lymphomas, and outcomes depend on both lymphoma control and the underlying immune condition.
Epstein–Barr virus contributes directly to some large B-cell lymphomas. The virus can persist in B cells and express proteins that promote survival and interfere with immune recognition. EBV-positive DLBCL occurs outside the transplant setting and is not confined to elderly people. Other virus-associated large-cell malignancies, such as those related to human herpesvirus 8, have their own diagnostic definitions and should not be merged into DLBCL, NOS.
Certain autoimmune diseases, including Sjögren syndrome and systemic lupus erythematosus, are associated with increased lymphoma risk. In rheumatoid arthritis, sustained inflammatory activity is an important part of the association; separating the effects of the disease from those of immunosuppressive treatment can be difficult. DLBCL can also arise from a pre-existing indolent lymphoma after years of evolution. Neither these associations nor its viral relationships make DLBCL contagious: malignant lymphoma cells are not ordinarily transmitted between people through everyday contact.
Cellular origins and molecular mechanisms
Normal B cells responding to an antigen may enter germinal centers, specialized structures in lymphoid tissue where antibody genes undergo mutation and selection. These processes generate effective immune responses but also create opportunities for errors in DNA repair. Chromosomal rearrangements and other acquired alterations can place growth-promoting genes under inappropriate control or disable pathways that normally remove damaged cells. DLBCL commonly contains multiple cooperating abnormalities rather than one universal initiating lesion.
Gene-expression profiling established two major cell-of-origin groups: germinal-center B-cell-like, or GCB, and activated B-cell-like, or ABC. A smaller proportion cannot be confidently assigned to either group by these methods. GCB tumors retain expression patterns resembling germinal-center B cells, whereas ABC tumors resemble cells undergoing later activation and differentiation. Immunohistochemical algorithms using markers such as CD10, BCL6, and MUM1 provide a practical approximation, often reported as GCB versus non-GCB. Non-GCB classification by staining is not identical to molecularly defined ABC disease, an important distinction when interpreting research.
Several recurring pathways illustrate the biological diversity. BCL2 promotes cell survival by inhibiting apoptosis, the regulated process of cell death. MYC controls programs involved in cellular growth and metabolism. BCL6 normally helps organize the germinal-center response but can contribute to malignancy when its regulation is disrupted. Some ABC tumors depend on persistent B-cell receptor signaling and activation of the NF-κB pathway. Alterations involving genes such as CD79B, MYD88, CARD11, and negative regulators of signaling can sustain these survival circuits.
A double-expressor tumor has increased MYC and BCL2 protein expression by immunohistochemistry. This is not equivalent to a double-hit lymphoma, in which chromosomal rearrangements are demonstrated by genetic testing. Protein overexpression can occur without the corresponding rearrangements, and rearranged tumors do not all have identical morphology or behavior. Keeping these concepts separate prevents a common diagnostic and prognostic misunderstanding.
Sequencing studies have identified genetic subgroups that cut across the older cell-of-origin categories. Clusters enriched for combinations such as MYD88 and CD79B alterations, BCL2 rearrangements with chromatin-regulator mutations, or NOTCH2-associated changes show distinct biological patterns. Classification systems developed by different research groups overlap but are not fully interchangeable. The surrounding tumor microenvironment also matters: T cells, macrophages, stromal cells, and immune-regulatory signals can support or restrain tumor growth. These findings explain why a visually similar group of tumors may respond differently to treatment, while also showing why a single biomarker rarely captures the whole disease.
Clinical presentation and tissue diagnosis
DLBCL commonly presents as a rapidly enlarging, usually painless lymph node or mass developing over weeks to months. Nodes in the neck, chest, abdomen, or groin may be involved. Extranodal disease is frequent and can affect the stomach, intestine, bone, skin, testes, or other organs. Symptoms depend on location: an abdominal mass may cause fullness or pain, thoracic disease may cause cough or breathlessness, and bone involvement may cause localized discomfort. Some patients have few symptoms despite extensive disease.
Systemic “B symptoms” are specifically defined as unexplained fever above 38°C, drenching night sweats, or unintentional loss of more than 10% of body weight over six months. Fatigue and appetite loss are also common but do not themselves meet that definition. Rapid growth can occasionally produce urgent complications, including airway compression, spinal cord compression, obstruction of an organ, or metabolic disturbances. Involvement of the bone marrow may contribute to low blood counts, although abnormal counts have many other possible causes.
A definitive diagnosis usually requires a tissue biopsy. An excisional or incisional sample preserves architecture and provides material for multiple tests; an adequately sized core biopsy can also establish the diagnosis in suitable circumstances. Fine-needle aspiration alone often cannot provide the architectural and molecular information needed to classify a suspected lymphoma. Examination by a pathologist with expertise in blood and lymphoid disorders integrates the clinical setting with the tissue findings.
On histopathology, diffuse sheets of large atypical lymphoid cells replace normal tissue. Immunohistochemical staining generally demonstrates B-cell markers such as CD20, PAX5, and CD79a, although expression varies and exceptions occur. Additional markers assess differentiation, proliferation, and alternative diagnoses. Ki-67 estimates the fraction of cells in the active cell cycle; a high value indicates rapid proliferation but is not sufficient by itself to define DLBCL or select a treatment. Flow cytometry can provide complementary evidence of an abnormal B-cell population.
Fluorescence in situ hybridization can detect rearrangements involving MYC, BCL2, and BCL6. Testing for EBV-encoded RNA identifies viral association when relevant. The differential diagnosis includes Burkitt lymphoma, other high-grade B-cell lymphomas, blastoid mantle cell lymphoma, lymphoblastic neoplasms, some forms of Hodgkin lymphoma, and non-lymphoid cancers with a similar appearance. Clinical location may further redirect classification toward a specific large B-cell entity. Neither imaging nor a blood test can replace this integrated tissue-based diagnosis.
Staging, response assessment, and prognostic models
After diagnosis, evaluation determines disease distribution, organ function, and factors relevant to treatment. Baseline studies commonly include a complete blood count, kidney and liver chemistry, lactate dehydrogenase, and testing for infections that may affect treatment safety. Lactate dehydrogenase is a nonspecific marker: an elevated concentration may reflect tumor burden or tissue injury, but it neither establishes lymphoma nor measures its extent precisely. Cardiac assessment is relevant because commonly used regimens contain an anthracycline.
Fluorodeoxyglucose positron emission tomography combined with computed tomography, usually called FDG PET/CT, is central to staging most DLBCL. It identifies metabolically active disease in lymph nodes and extranodal sites and provides a baseline for response assessment. Bone marrow biopsy is no longer automatically required in every patient when high-quality PET/CT is available, although it can provide useful information when discordant marrow disease or another hematologic condition is suspected. Neurologic symptoms may lead to magnetic resonance imaging or cerebrospinal fluid evaluation; these are not routine tests for every case.
The Lugano framework uses anatomic stages derived from the Ann Arbor system. Stage I generally denotes a single lymph-node region or a single localized extranodal site. Stage II involves multiple nodal regions on the same side of the diaphragm; stage III involves nodes on both sides. Stage IV describes diffuse or disseminated extranodal involvement, such as multifocal marrow or liver disease. Bulky masses are recorded separately, with size thresholds varying between protocols. Unlike many solid tumors, stage IV DLBCL does not necessarily imply that cure is unlikely.
The International Prognostic Index, introduced in 1993, assigns one adverse factor each for age over 60 years, elevated lactate dehydrogenase, impaired performance status, stage III or IV disease, and involvement of more than one extranodal site. The age-adjusted IPI and later models refine risk estimation in particular populations or treatment eras. These scores describe outcomes in groups, not certainties for individuals. The CNS-IPI adds kidney or adrenal involvement to the standard IPI factors to estimate the risk of central nervous system recurrence.
Treatment response is commonly assessed with the five-point Deauville scale, which compares residual FDG uptake with mediastinal blood-pool and liver activity. At the end of standard treatment, scores of 1–3 generally indicate complete metabolic response. Residual uptake can also arise from inflammation or infection, so a positive scan does not invariably establish viable lymphoma. Interim PET findings have prognostic associations, but changing treatment solely because of an interim result has not consistently improved outcomes across clinical settings.
Initial treatment and the evidence behind it
Initial therapy is usually systemic because DLBCL can involve microscopic sites beyond an apparent mass. The long-established reference regimen is R-CHOP: rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone. Rituximab is a monoclonal antibody directed against CD20, a surface molecule present on most DLBCL cells. Its actions include recruiting immune mechanisms that destroy antibody-coated B cells. The chemotherapy components damage DNA, interfere with cell division, or alter lymphoid-cell survival, while prednisone is a corticosteroid with direct lympholytic effects.
Randomized studies established that adding rituximab to CHOP improved disease control and survival. A landmark trial published in 2002 demonstrated this benefit in previously untreated patients aged 60–80 years. Subsequent studies extended the evidence across other populations. Six cycles of R-CHOP became a common approach for advanced disease, while selected limited-stage presentations can be treated with shorter systemic courses, sometimes combined with involved-site radiation therapy. The exact approach depends on stage, bulk, prognostic features, response assessment, and the evidence applicable to the patient group.
Polatuzumab vedotin is an antibody–drug conjugate targeting CD79b that delivers a microtubule-disrupting payload. In the POLARIX trial, published in 2022, polatuzumab replaced vincristine in a regimen termed pola-R-CHP. Among previously untreated patients with intermediate- or high-risk disease as defined by the trial, two-year progression-free survival was approximately 76.7%, compared with 70.2% for R-CHOP. The original analysis did not demonstrate a significant overall-survival difference. The study established pola-R-CHP as another first-line option in jurisdictions and patient groups covered by its authorization, rather than proving that every DLBCL presentation requires the same regimen.
More intensive treatment is not automatically more effective. In the randomized Alliance/CALGB 50303 study, dose-adjusted EPOCH-R did not improve progression-free or overall survival over R-CHOP in the overall study population and caused greater toxicity. That result does not resolve every question involving rare molecular subgroups, which were not all adequately represented. Likewise, evidence from DLBCL, NOS cannot simply be transferred to primary mediastinal, primary central nervous system, or other separately classified large B-cell lymphomas.
Treatment delivery also depends on physiological reserve. Very old or frail patients may receive attenuated regimens, while significant cardiac disease may preclude standard anthracycline exposure. Surgery is generally used to obtain tissue or address complications, not as the primary curative treatment for systemic DLBCL. These distinctions reflect the disease’s systemic biology and the balance between lymphoma control and treatment-related harm.
Relapsed and refractory disease
“Refractory” disease fails to achieve an adequate response or progresses during initial treatment; “relapsed” disease returns after a response. These categories include biologically different situations. Progression during therapy or soon afterward generally indicates greater resistance than recurrence several years later. Re-biopsy can establish whether a new lesion is recurrent DLBCL, another lymphoma, infection, or an unrelated process, and can identify changes relevant to subsequent therapy. Treatment options depend on timing, prior exposure, organ function, performance status, and access to specialized care.
For decades, the principal potentially curative strategy for fit patients with recurrent, chemotherapy-sensitive disease was salvage chemotherapy followed by high-dose treatment and autologous hematopoietic stem-cell transplantation. Regimens such as R-ICE and R-DHAP were used to obtain disease control before transplantation. Autologous stem cells restore blood formation after marrow-toxic therapy; they do not provide the donor-derived immune effect associated with allogeneic transplantation. This approach remains important, particularly for some patients with later relapse.
Chimeric antigen receptor T-cell therapy changed the treatment pathway for early relapsed or refractory large B-cell lymphoma. The patient’s T cells are collected, genetically modified to recognize a tumor antigen—usually CD19—and expanded before infusion after lymphodepleting chemotherapy. The ZUMA-7 and TRANSFORM randomized trials demonstrated improved event-free survival with axicabtagene ciloleucel and lisocabtagene maraleucel, respectively, compared with conventional second-line strategies in their studied populations, which included primary refractory disease or relapse within 12 months of first-line therapy. A separate trial, BELINDA, did not show the same benefit for tisagenlecleucel in its particular second-line setting, illustrating that products and trial pathways are not interchangeable.
Other immune and targeted treatments broaden options after multiple previous therapies or when transplantation is unsuitable. CD20×CD3 bispecific antibodies, including glofitamab and epcoritamab, connect malignant B cells with T cells and stimulate immune killing. Antibody–drug conjugates include polatuzumab vedotin and the CD19-directed agent loncastuximab tesirine. Tafasitamab, a CD19-directed antibody, has been used with lenalidomide in defined relapsed or refractory settings. Authorizations and preferred sequencing differ between health systems and continue to evolve.
These treatments have distinct limitations. CAR T-cell therapy requires specialized manufacturing and clinical infrastructure, and rapidly progressing disease may require interim treatment. Bispecific antibodies offer a different delivery model but can also cause immune-mediated toxicity. Antigen loss, impaired T-cell function, tumor microenvironment changes, and intrinsic tumor resistance can produce failure after immune therapy. Allogeneic transplantation remains a specialized option for selected patients, with the potential benefit of donor immunity balanced against substantial treatment-related risks.
Treatment complications and supportive care
DLBCL treatment can produce complications through direct tissue injury, suppression of blood formation, immune disruption, and rapid tumor destruction. The pattern depends on the regimen and the patient’s underlying health. Cytotoxic therapy commonly causes neutropenia, anemia, thrombocytopenia, nausea, mucosal injury, and hair loss. Neutropenic infection can progress to sepsis, sometimes with few localizing symptoms because inflammatory responses are impaired. Growth-factor support, infection assessment, and antimicrobial strategies are therefore integrated into many treatment protocols.
Tumor lysis syndrome results when large numbers of malignant cells release potassium, phosphate, and nucleic-acid breakdown products into the circulation. It can cause acute kidney injury, cardiac rhythm disturbances, and seizures. Risk is influenced by tumor burden, proliferation rate, kidney function, baseline biochemical abnormalities, and treatment sensitivity. Clinical prevention protocols use risk-adapted monitoring, hydration, and uric-acid-lowering medicines. Tumor lysis can occasionally occur before treatment in rapidly proliferating disease.
Individual drugs have characteristic toxicities. Doxorubicin can damage heart muscle, with risk affected by cumulative exposure and pre-existing cardiovascular conditions. Vincristine can cause peripheral neuropathy and autonomic dysfunction, including constipation. Corticosteroids can contribute to hyperglycemia, sleep disruption, muscle weakness, and infection susceptibility. Rituximab can cause infusion reactions and prolonged B-cell depletion. Because anti-CD20 treatment can reactivate hepatitis B, pretreatment viral assessment and risk-based antiviral management are established elements of care.
Immune-effector therapies require additional surveillance. Cytokine release syndrome may follow CAR T-cell therapy or bispecific-antibody treatment and ranges from fever to hypotension and respiratory compromise. Immune effector cell-associated neurotoxicity syndrome can include impaired attention, language disturbance, confusion, or seizures. Management follows therapy-specific grading systems and protocols. Prolonged cytopenias, low immunoglobulin levels, and infections can persist after cellular therapy, so toxicity assessment extends beyond the infusion period.
Supportive care also addresses effects not captured by tumor-response measures. Gonadal injury can affect fertility, while prolonged treatment may impair physical function, nutrition, employment, and emotional well-being. Rehabilitation, symptom control, transfusion support, and psychosocial services form part of multidisciplinary lymphoma care. Palliative care can accompany disease-directed treatment and is not limited to the final phase of life. Long-term follow-up evaluates both recurrence and late effects, including selected cardiovascular complications and second malignancies, whose risks vary with age, exposures, and length of survival.
Outcomes, survivorship, and current research
DLBCL is frequently curable. With modern first-line immunochemotherapy, roughly 60–70% of patients achieve durable disease control, although estimates differ according to age, risk distribution, biological subtype, treatment, and the definition of cure. An important distinction is between overall survival, which counts death from any cause, and progression-free or event-free survival, which also incorporates specified disease-related events. These measures cannot be substituted for one another when comparing trials or population registries.
Most treatment failures occur relatively early. Patients who remain event-free for two years after diagnosis have substantially more favorable subsequent survival than patients assessed at diagnosis, and several cohorts have found survival approaching that of age- and sex-matched general populations. This is a group-level observation rather than a guarantee. Late relapse still occurs, sometimes involving an underlying indolent lymphoma rather than recurrent aggressive disease. New symptoms or lesions therefore require interpretation in the context of the original pathology and treatment history.
Survivorship is shaped by more than recurrence risk. Older survivors may have competing illnesses, while younger survivors may live for decades with the consequences of prior treatment. Persistent fatigue, neuropathy, impaired physical capacity, anxiety about recurrence, and financial disruption can affect quality of life. Routine surveillance imaging has not consistently demonstrated a survival benefit in asymptomatic patients in complete remission and can generate false-positive findings, additional biopsies, radiation exposure, and anxiety. Follow-up strategies consequently emphasize clinical assessment, with imaging used according to clinical circumstances rather than as an indefinite automatic schedule.
One major research direction is the measurement of circulating tumor DNA. Fragments released by lymphoma cells into blood can carry tumor-specific mutations or immunoglobulin sequences. Highly sensitive assays may detect molecular disease before conventional imaging shows recurrence and may quantify early treatment response. Important unresolved issues include assay standardization, thresholds for meaningful positivity, false-negative results in low-shedding disease, and whether changing treatment on the basis of a molecular result improves outcomes. A prognostic association alone does not establish the benefit of an intervention.
Another direction is biologically selected treatment. Trials are testing whether genetic subgroups, signaling dependencies, immune characteristics, or combinations of these can identify patients who benefit from particular targeted agents. Studies also examine earlier use of cellular therapies and bispecific antibodies, strategies after CAR T-cell failure, and ways to reduce treatment intensity without sacrificing cure in favorable-risk disease. The central methodological challenge is to demonstrate benefit in adequately designed randomized controlled trials, rather than infer effectiveness solely from laboratory plausibility or high response rates in small, selected cohorts.