Residency · Residency · Hematology Thrombosis

Diffuse Large B-Cell Lymphoma

Introduction

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma, accounting for approximately 30 to 35% of all NHL cases. The incidence is approximately 7 per 100,000 per year, with a median age at diagnosis of 65 years. DLBCL is an aggressive lymphoma that is curable in approximately 60% of patients with first-line therapy. Prognosis is determined by the molecular subtype, the International Prognostic Index score, and the response to initial treatment. Molecular classification and the development of targeted therapeutic approaches are actively reshaping the treatment landscape.

Pathology and Molecular Classification

Cell of Origin (COO) by Gene Expression Profiling

Gene expression profiling has identified two principal cell-of-origin subtypes of DLBCL. Germinal center B-cell (GCB) DLBCL accounts for approximately 45 to 50% of cases, carries a better prognosis, and is characterized by BCL2 rearrangement, EZH2 mutations, and CREBBP mutations. Activated B-cell (ABC, also known as non-GCB) DLBCL accounts for 30 to 35% of cases, carries a worse prognosis, and is driven by NF-kappa-B pathway activation with frequent MYD88 and CD79B mutations and BCL2 amplification. Approximately 10 to 20% of cases remain unclassifiable. The gold standard for cell-of-origin determination is gene expression profiling using platforms such as the Lymph2Cx NanoString assay. When gene expression profiling is unavailable, immunohistochemical algorithms (such as the Hans algorithm using CD10, BCL6, and MUM1) provide approximately 85% concordance with gene expression profiling.

Molecular Subtypes (Chapuy/Schmitz Classification)

More granular molecular classification has identified biologically distinct subtypes. The MCD subtype (MYD88/CD79B mutated) is an ABC-derived subset with a characteristic predilection for extranodal sites including the CNS, testes, and breast, and may respond to BTK inhibitor therapy. The BN2 subtype (BCL6/NOTCH2 mutated) has marginal zone-like features and an intermediate prognosis. The N1 subtype (NOTCH1 mutated) includes T-cell-rich and histiocyte-rich cases with a poor prognosis. The EZB subtype (EZH2/BCL2) is a GCB-derived subset with BCL2 rearrangement and EZH2 mutations that may respond to EZH2 inhibitor therapy. The ST2 subtype (SGK1/TET2) is a favorable GCB-derived subset. The A53 subtype (aneuploidy/TP53) carries TP53 mutations or deletions and a very poor prognosis.

High-Grade B-Cell Lymphoma

Double-hit lymphoma is defined by the co-occurrence of MYC rearrangement with BCL2 rearrangement (the most common combination) or BCL6 rearrangement. Triple-hit lymphoma involves rearrangements of all three genes. These rearrangements must be identified by FISH and carry an adverse prognosis when treated with standard R-CHOP. The WHO 5th Edition designates these cases as "High-grade B-cell lymphoma with MYC and BCL2 rearrangements." Double-expressor lymphoma (DEL), defined by MYC protein expression of 40% or more and BCL2 protein expression of 50% or more by immunohistochemistry without underlying gene rearrangement, is found in approximately 30% of DLBCL and carries an intermediate adverse prognosis.

Risk Stratification

International Prognostic Index (IPI)

The IPI incorporates five adverse prognostic factors: age above 60, LDH above the upper limit of normal, ECOG performance status of 2 or greater, Ann Arbor stage III-IV, and more than 1 extranodal site of involvement. Risk groups include low (0-1 factors), low-intermediate (2), high-intermediate (3), and high (4-5). The 5-year overall survival with R-CHOP ranges from 91% in the low-risk group to 59% in the high-risk group.

IPI factors: age >60, LDH > ULN, ECOG PS ≥2, stage III-IV, >1 extranodal site.

CNS-IPI

The CNS-IPI adds renal or adrenal involvement to the standard IPI factors to predict the risk of CNS relapse. A high-risk CNS-IPI score (4-6 points) is associated with a CNS relapse rate of approximately 12% at 2 years and constitutes an indication for CNS prophylaxis.

<image>A molecular classification diagram of DLBCL. Show the cell-of-origin classification (GCB vs. ABC) determined by gene expression profiling as two main branches. Within each branch, display the molecular subtypes: GCB → EZB (EZH2, BCL2 translocation), ST2 (SGK1, TET2), and A53 (TP53); ABC → MCD (MYD88, CD79B), BN2 (BCL6, NOTCH2), N1 (NOTCH1). For each subtype, list: frequency, key genetic alterations, typical clinical features, prognosis (shown as a color gradient from green/favorable to red/poor), and potential targeted therapies (BTK inhibitors for MCD, EZH2 inhibitors for EZB). Include a separate box highlighting double-hit lymphoma (MYC + BCL2 rearrangement by FISH) with its relationship to the EZB subtype. Show the Hans IHC algorithm (CD10, BCL6, MUM1) as a simplified decision tree for COO determination when GEP is unavailable. Medical education style with clean layout.</image>

First-Line Treatment

R-CHOP (Standard)

R-CHOP remains the standard first-line regimen for DLBCL, consisting of rituximab at 375 mg/m2, cyclophosphamide at 750 mg/m2, doxorubicin at 50 mg/m2, vincristine at 1.4 mg/m2 (capped at 2 mg), and prednisone at 100 mg daily for days 1 through 5. Six cycles are administered at 21-day intervals. This regimen achieves a complete response rate of 75 to 80% and a 5-year overall survival of 60 to 65%. G-CSF support is recommended per NCCN guidelines for patients aged 65 and older or those with significant comorbidities.

Pola-R-CHP

Polatuzumab vedotin, an anti-CD79b antibody-drug conjugate, replaces vincristine in the R-CHOP backbone to create the Pola-R-CHP regimen. The POLARIX trial demonstrated a 2-year PFS of 76.7% with Pola-R-CHP compared to 70.2% with R-CHOP (hazard ratio 0.73, p=0.02), though no overall survival difference has been detected to date. Pola-R-CHP received FDA approval in 2023 as first-line treatment for DLBCL and is the NCCN-preferred regimen for patients with an IPI score of 2 or greater. The peripheral neuropathy profile is similar to R-CHOP, and neutropenia requires monitoring.

Double-Hit Lymphoma Treatment

Standard R-CHOP is inadequate for double-hit lymphoma, consistently producing inferior PFS and OS compared to more intensive regimens. DA-EPOCH-R (dose-adjusted etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) has been widely adopted as the preferred regimen for double-hit lymphoma, with 4-year event-free survival of approximately 70% in prospective studies. Six cycles are administered with dose adjustment based on the absolute neutrophil count nadir, following a complex pharmacokinetic dosing algorithm. The ALLIANCE/CALGB 50303 trial showed no difference between DA-EPOCH-R and R-CHOP for unselected DLBCL, confirming that the benefit of DA-EPOCH-R is concentrated in the double-hit and triple-hit subsets. Data on Pola-R-CHP for double-hit lymphoma remain insufficient, and DA-EPOCH-R is generally preferred.

Limited Stage (I-II) Disease

For stage I-II non-bulky disease, options include 3 to 4 cycles of R-CHOP plus involved-site radiation therapy, or 6 cycles of R-CHOP without radiation using a PET-adapted approach. The FLYER trial demonstrated non-inferiority of 4 cycles of rituximab plus 4 cycles of CHOP compared to 6 cycles of R-CHOP in young, low-risk, stage I-II patients. The LYSA/GOELAMS PET-adapted strategy administers 4 cycles of R-CHOP followed by PET assessment; PET-negative patients receive 2 additional cycles of R-CHOP without radiation.

CNS Prophylaxis

CNS prophylaxis is indicated for patients with high-risk anatomic sites (testicular, breast, kidney or adrenal, uterine, or epidural involvement), high-risk CNS-IPI scores (4-6), and double-hit lymphoma with CNS-IPI of 3 or greater. Prophylactic options include intrathecal methotrexate at 12 mg administered with each chemotherapy cycle and high-dose intravenous methotrexate at 3 to 3.5 g/m2 interdigitated between R-CHOP cycles. There is growing consensus that intrathecal methotrexate alone may be insufficient for adequate CNS prophylaxis and that high-dose systemic methotrexate provides more reliable CSF penetration, though it is logistically more challenging to administer.

Response Assessment

Interim PET-CT after 2 to 4 cycles of therapy is prognostic but does not currently guide treatment changes in DLBCL outside of clinical trials. End-of-treatment PET-CT is essential: Deauville scores 1 through 3 indicate a complete metabolic response, while scores 4 or 5 require biopsy to confirm residual viable disease before initiating salvage therapy. Post-treatment surveillance consists of clinical examination and laboratory studies every 3 to 6 months for 2 years; routine surveillance PET scanning is not recommended due to a high false-positive rate.

Relapsed/Refractory DLBCL

CAR T-Cell Therapy (Second-Line)

CAR T-cell therapy has fundamentally changed the management of relapsed DLBCL. Axicabtagene ciloleucel (Yescarta), evaluated in the ZUMA-7 trial, demonstrated superior 4-year event-free survival (41.8% versus 24.0%, hazard ratio 0.73) and 4-year overall survival (54.2% versus 46.0%) compared to standard salvage chemotherapy followed by autologous HSCT in patients with primary refractory or early relapse (within 12 months) DLBCL. Lisocabtagene maraleucel (Breyanzi), evaluated in the TRANSFORM trial, demonstrated similarly superior EFS compared to the standard salvage-HSCT pathway. CAR T-cell therapy is now the preferred second-line treatment for primary refractory or early relapse DLBCL, supplanting salvage chemotherapy plus autologous HSCT in this population. In the third-line and beyond setting, ZUMA-1 (axi-cel: ORR 83%, CR 58%) and TRANSCEND (liso-cel: ORR 73%, CR 53%) demonstrated meaningful response rates. CRS and ICANS are managed with tocilizumab and dexamethasone, respectively.

Bispecific Antibodies

Bispecific antibodies represent an important advance for relapsed/refractory DLBCL, offering "off-the-shelf" availability without the manufacturing delays inherent to CAR T-cell therapy. Glofitamab, a CD20 by CD3 bispecific antibody, achieved an overall response rate of 52% and a complete response rate of 39% in the NP30179 trial in patients with relapsed/refractory DLBCL after 2 or more prior lines. Epcoritamab, a subcutaneously administered CD20 by CD3 bispecific antibody, demonstrated an ORR of 63% and CR of 39% in the EPCORE NHL-1 trial. Both agents employ step-up dosing to mitigate CRS risk, and glofitamab requires obinutuzumab pretreatment.

Salvage Chemotherapy + Auto-HSCT

Standard salvage chemotherapy followed by autologous HSCT remains appropriate for patients with late relapse (beyond 12 months) who demonstrate chemosensitive disease. Common salvage regimens include R-ICE, R-DHAP, and R-GDP. PET-negative status before autologous HSCT predicts a better outcome.

Polatuzumab + BR (Tafasitamab + Lenalidomide)

Polatuzumab vedotin combined with bendamustine and rituximab achieved an ORR of 45% and CR of 40% in relapsed/refractory DLBCL in the ROMULUS study. Tafasitamab, an anti-CD19 antibody, combined with lenalidomide, achieved a CR rate of 43% in transplant-ineligible relapsed/refractory DLBCL in the L-MIND trial.

<image>A treatment sequence diagram for DLBCL showing the evolving treatment landscape from first-line through multiple relapses. First-line: Pola-R-CHP or R-CHOP x 6 (or DA-EPOCH-R for double-hit). Show response assessment with end-of-treatment PET. For primary refractory or early relapse (<12 months): CAR T-cell therapy (axi-cel or liso-cel) as preferred approach (ZUMA-7/TRANSFORM data). For late relapse (>12 months): salvage chemo (R-ICE/R-DHAP) → auto-HSCT if chemosensitive. For post-CAR T relapse or CAR T-ineligible: bispecific antibodies (glofitamab, epcoritamab), polatuzumab-BR, tafasitamab-lenalidomide, or clinical trials. Show response rates for each treatment modality. Include a timeline showing the shift from auto-HSCT to CAR T as second-line standard for early relapse. Modern clinical pathway format with response rate annotations.</image>

Key Clinical Pearls

• FISH for MYC, BCL2, and BCL6 rearrangements should be performed on ALL newly diagnosed DLBCL; missing a double-hit lymphoma leads to inadequate therapy with R-CHOP
• Pola-R-CHP is now a preferred first-line option for DLBCL with IPI >=2 based on POLARIX; it replaces vincristine (not added on top of R-CHOP)
• CAR T-cell therapy has moved to second-line for primary refractory/early relapse DLBCL, supplanting salvage chemotherapy + auto-HSCT in this population
• Bispecific antibodies (glofitamab, epcoritamab) provide off-the-shelf options for r/r DLBCL without the manufacturing delay of CAR T
• CNS prophylaxis with high-dose IV methotrexate (3-3.5 g/m2) is more effective than intrathecal methotrexate; target high-risk patients (testicular, breast, kidney, CNS-IPI >=4, double-hit)
• MYD88 L265P + CD79B mutations (MCD subtype) identify patients who may benefit from BTK inhibitor addition; multiple trials ongoing

References

1. Tilly H, et al. Diffuse large B-cell lymphoma: ESMO Clinical Practice Guidelines. Ann Oncol. 2015;26(Suppl 5):v116-v125.
2. Tilly H, et al. Polatuzumab vedotin in previously untreated diffuse large B-cell lymphoma (POLARIX). N Engl J Med. 2022;386(4):351-363.
3. Locke FL, et al. Axicabtagene ciloleucel as second-line therapy for large B-cell lymphoma (ZUMA-7). N Engl J Med. 2022;386(7):640-654.
4. Dickinson MJ, et al. Glofitamab for relapsed or refractory diffuse large B-cell lymphoma. N Engl J Med. 2022;387(24):2220-2231.
5. Schmitz R, et al. Genetics and pathogenesis of diffuse large B-cell lymphoma. N Engl J Med. 2018;378(15):1396-1407.

Read this lecture as Markdown