Introduction

Diffuse large B-cell lymphoma (DLBCL) is the most common lymphoma encountered throughout the world with significant geographical variation. There were an estimated 69,740 cases of non-Hodgkin’s lymphoma (NHL) in the United States in 2013, of which approximately 25,000 cases were some form of DLBCL. Pathologically, DLBCL is a heterogeneous disorder composed of several entities within the World Health Organization (WHO) classification (Table 40.1). This chapter reviews the pathobiology of DLBCL as well as consideration regarding staging, management, and prognosis.

Multiple Choice and Discussion Questions

1. The incidence of DLBCL has steadily declined over the course of the last 10 years.
   
2. Systemic lupus erythematosus (SLE) can be associated with an increased risk of DLBCL.
   
3. Familial history of NHL increases the risk for the development of NHL of any histology by 1.8-fold.

The incidence of NHL doubled in the United States and other developed countries between 1975–1977 and 2004–2006 (SEER Cancer Statistics Review, 1975–2006). Multiple factors have contributed to the rise in incidence, including the rise in AIDS-related lymphoma, improved diagnosis, and an aging population; however, these factors do not fully account for the rise in incidence. Factors associated with an increased risk of NHL include occupation exposure to agricultural chemicals, immunodeficiency associated with HIV infection, immune suppression associated with solid-organ and allogeneic bone marrow transplantation, and prior chemotherapy. Viral agents associated with the development of lymphoma include the herpes viruses Epstein–Barr virus (EBV) and human herpes virus 8 (HHV8), the retrovirus human T-lymphotropic virus (HTLV1), and hepatitis B virus (HBV) and hepatitis C virus (HPC) infections. Autoimmune diseases, particularly Sjögren syndrome and SLE, are associated with increased risk of NHL, including DLBCL. A familial history of NHL increases the risk for the development of NHL of any histology by 1.8-fold and increases the risk of DLBCL specifically by 2.3-fold.

The foundation for optimal management of DLBCL lies in establishing an accurate histopathologic diagnosis. In the absence of specialized hematopathology expertise, pathologic consultation is often necessary. Classification of malignant lymphoma has evolved as pathologists and clinicians have gained greater insight into the immunobiology and, more recently, molecular pathology of individual entities. The current standard for classification is the 2008 WHO classification of lymphoid neoplasms; however, it is recognized that classification of lymphoma remains a work in progress. Recent series have demonstrated that 6% (in a series restricted to DLBCL and follicular lymphoma) to 18% (in a series of all lymphomas) of diagnoses are changed on review at comprehensive cancer centers in a way that would likely impact clinical management. Fine-needle aspirate is not acceptable for an initial diagnosis of DLBCL. Ideally, the patient should undergo an excisional or incisional lymph node biopsy. Laparoscopic-guided intra-abdominal biopsies have been shown to be more effective and minimally more difficult than computed tomography (CT)-guided core needle biopsy.

1. True
   
2. False

DLBCL is a heterogeneous disorder; even the most common subtype, DLBCL-NOS, is itself heterogeneous. GEP has demonstrated that DLBCL-NOS represents at least two different diseases based on the cell of origin (COO): germinal center DLBCL (GC-DLBCL) and activated B-cell DLBCL (ABC-DLBCL). GC-DLBCL and ABC-DLBCL have distinct oncogenic pathways. GC-DLBCL is characterized by genomic instability caused by PTEN deletion, ING1 deletion, MDM2 gain or amplification, TP53 mutation, mTOR activation as a consequence of miR-17-92 amplification, and BCL2 activation by translocation. ABC-DLBCL is characterized by recurrent mutations in MYOM2, TNFAIP3, CD79B, PRDM1, and CARD11; deletion of INK4A-ARF; trisomy 3; and 19p gain or amplification, resulting in activation of nuclear factor kappa B (NF-kB). Recurrent mutations seen in more than 10% of cases of both GC-DLBCL and ABC-DLBCL include MLL2, CREBBP, TP53, CD36, B2M, and MEF2B. Both progression-free survival (PFS) and overall survival (OS) are superior for GC-DLBCL compared to ABC-DLBCL following treatment with standard R-CHOP chemotherapy (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone); at 4 years, the PFS and OS differences are approximately 40% and 30%, respectively (also reviewed in Chapter 52). Though COO has clear prognostic implications, routine clinical application has proven to be difficult. The gold standard for determination of COO is GEP. New technologies, includ­ing quantitative nuclease protection (High Throughput Genomics, Tucson, AZ), microfluidics quantitative polymerase chain reaction (Fluidigm, South San Francisco, CA) and direct assay of RNA with capture and detection probes (Nanostring, Seattle, WA), have made it possible to perform GEP on formalin-fixed paraffin-embedded (FFPE) tissue; however, they are not in routine clinical use for determination of COO.

1. Yes
   
2. No

There have been attempts to determine COO by IHC. The most widely used algorithm includes three antigens: CD10, IRF4–MUM1, and BCL6. In this algorithm, CD10 expression is associated with GC-DLBCL, regardless of expression of other antigens. In the absence of CD10 expression, IRF4–MUM1 expression is associated with non-GC-DLBCL. In cases negative for both CD10 and IFR4–MUM1, BCL6 expression correlates with GC-DLBCL and lack of expression with non-GC-DLBCL. This has been the most commonly used algorithm; however, the ability to predict outcome in different patient cohorts has been variable. The alternative IHC-based algorithms Colomo (IRF4–MUM1, CD10, and BCL6), Muris (BCL2, CD10, and IRF4–MUM1), Choi (GCET1, IRF4–MUM1, CD10, FOXP1, and BCL6), and Tally (CD10, GCET1, IRF4–MUM1, FOXP1, and LMO2) have been reported to predict the outcome of patients. Both Choi and Tally have reported greater specificity than the Hans algorithm in classification of GC- and non-GC-DLBCL. The ability of these newer algorithms to predict outcome by COO has also been called into question. It appears that IHC algorithms can enrich for patients with either GC- or ABC-COO but are not highly reproducible in predicting for PFS and OS.

Another increasingly recognized challenge has been the management of cases of DH or TH lymphoma. DH cases were originally described as having translocation t(14;18)(q32 q21.3) involving both IGH–BCL2 and MYC–8q24, and TH cases have additional translocation of BCL6–3q27. Some authors have included cases with MYC and BCL6 rearrangements in the definition of DH lymphoma. BCL2–MYC DH lymphomas account for 1–8% of cases, and TH lymphomas are rare, occurring in 0–3% of cases is various series. DH and TH lymphomas have a poor prognosis when treated with conventional R-CHOP chemotherapy (discussed in Chapters 45 and 46). DH and TH lymphomas as defined above are restricted to patients with GC-DLBCL. More recently, concurrent overexpression of MYC (≥40% of cells) and BCL2 (≥70% of cells) proteins (IHC DH) has been reported to be associated with a poor prognosis with a similar magnitude of impact as the genetic DH lymphomas; IHC DH lymphomas are independent of COO. Optimal treatment regimens for DH, TH, and IHC DH lymphomas have not yet been defined.

Lumbar puncture with examination of CSF by flow cytometry is indicated in patients with ≥2 extranodal sites, elevated lactate dehydrogenase (LDH), and involvement of paranasal sinus, testes, or epidural sites, or bone marrow involvement with large cell.

Patient planned to receive rituximab should have a hepatitis B surface antigen assessment given the risk of viral reactivation. Universal screening for hepatitis B surface antigen was found to be cost-effective in patients receiving R-CHOP chemotherapy. There is some controversy about the added benefit of hepatitis B core antibody testing. Approximately 10% of patients will be hepatitis B core antibody positive, and the risk of reactivation is approximately 4%. In patients at risk for hepatitis B reactivation, prophylaxis with an antiviral is superior to treatment upon activation. Recent data suggest that entecavir is superior to lamivudine for prophylaxis. The risk of reactivation persists for at least 6 months beyond the completion of chemotherapy, and prophylaxis should be continued during this period.