Childhood Lymphoma

Introduction

Substantial clinical and laboratory advances have improved our understanding of both the pathogenesis and treatment of the malignant lymphomas of childhood and adolescence, which comprise Hodgkin lymphoma (HL) and the non-Hodgkin lymphomas (NHLs). These are the third most common type of cancer in children in the United States, comprising approximately 15% of newly diagnosed cases in this age group each year. Among children younger than 15 years of age, the NHLs account for approximately 60% of cases. However, when children up to age 18 years are included, there is a slight predominance of HL.

The NHLs of childhood are markedly different from those of adulthood. Whereas diffuse high-grade extranodal subtypes account for the majority of pediatric cases, low- and intermediate-grade lymphomas are predominant in adults. These differences are probably due in part to age-related maturational changes in the immune system and consequently in the types of cells susceptible to malignant transformation. The differences between adults and children with respect to histologic subtype underlie the differing clinical features, staging, and treatment strategies in these age groups. The age-related differences in biology across the pediatric, adolescent, and young adult age groups and their respective impacts on treatment have been reviewed.

The clinical presentation, staging, histologic subtypes, and treatment strategies in children and adults with HL are less dissimilar; however, some differences are worth noting. Epidemiologic studies have suggested three distinct forms that depend on age: the childhood form in patients 14 years of age or younger, the young adult form (ages 15 to 34 years), and an older adult form in individuals 55 to 74 years of age. Among the four histologic subtypes of HL described in the Rye classification system, the nodular sclerosing subtype is most common in children, occurring in 40% of younger children and 70% of adolescents. Mixed-cellularity HL occurs in approximately 30% of cases and is more common in children with HIV infection and in those less than 10 years of age; it is frequently associated with an advanced stage (with extranodal extension) at presentation. Lymphocyte-predominant HL accounts for approximately 10% to 15% of pediatric cases, is usually associated with localized disease at presentation, and occurs more commonly in younger patients and in boys. The fourth subtype of HL, lymphocyte depletion, is very rare in children.

Despite excellent event-free survival (EFS) rates in HL, there are well-recognized challenges related to the sequelae of therapy, which include endocrine dysfunction, chemotherapy-induced sterility, radiation-induced abnormalities in bone growth, chemotherapy- and radiation-related second cancers, and late cardiac deaths. Most current studies are exploring strategies that maintain an excellent treatment result while reducing late effects. For example, a combined-modality approach with low-dose radiation and combination chemotherapy may reduce the bone growth abnormalities associated with high-dose extended-field radiation therapy. However, because therapy is modified in an attempt to reduce the risk of late effects, the rates of EFS may be compromised. For example, attempts at reducing alkylating agent–related late effects in patients with advanced-stage disease by substituting other chemotherapeutic agents have lowered EFS rates. There have been some recent clinical trials that have also attempted to reduce toxicity without compromising outcome. In this regard, among children with favorable-risk HL and a complete early response to vincristine, doxorubicin, methylprednisolone (VAMP) chemotherapy, a high rate of 2-year EFS could be achieved with limited use of radiotherapy. The results of the Children's Cancer Group (CCG) 5942 study, which examined the benefit of adding involved field radiation therapy (IFRT) for patients who achieved a complete response to chemotherapy, were analyzed and revealed a statistically significant improvement in EFS for those who received IFRT; however, there was no associated improvement in overall survival. The refinement of a risk-adapted treatment approach to the management of pediatric HL remains a challenge. Current studies are examining the benefits of incorporating noel immunotherapeutic agents into frontline therapy in an effort to both improve outcome and limit unwanted side effects of therapy. These and other issues regarding the management of children with HL are discussed in Chapter 102 . The remainder of this chapter focuses on the NHLs of childhood.

Epidemiology and Pathogenesis

Non-Hodgkin lymphomas may occur at any age in childhood but are unusual in children younger than age 3 years; the median age at diagnosis is approximately 10 years. There are approximately 500 to 750 new cases of pediatric NHL in the United States each year. In contrast to HL, which has a bimodal age distribution with peaks in early and late adulthood, the incidence of NHLs increases steadily with age. NHL occurs almost twice as commonly in whites as in blacks and two to three times more often in boys than in girls; the explanation for these differences has yet to be elucidated.

There are specific populations at increased risk for the development of NHL. These include individuals with primary immunodeficiency syndromes, including ataxia-telangiectasia (A-T), Wiskott-Aldrich syndrome, X-linked lymphoproliferative syndrome (XLP), common variable immunodeficiency, Nijmegen syndrome, and autoimmune lymphoproliferative syndrome (ALPS). More recently, an increased risk for the development of NHL has been reported to occur in children with constitutional mismatch repair deficiency syndrome (CMMRDS). It is important that these syndromes be recognized so that appropriate therapy can be designed. For example, in the management of children with A-T who develop a malignancy, IFRT and radiomimetics such as bleomycin should be avoided. Children with A-T are also at increased risk for the development of late-onset hemorrhagic cystitis after exposure to cyclophosphamide. Boys with XLP are at increased risk for developing fatal infectious mononucleosis, B-cell lymphomas, or both. XLP should be considered in any boy with a high-grade B-cell lymphoma whose brother has had either fatal infectious mononucleosis or B-cell lymphoma and in any boy who has had two primary B-cell lymphomas. Children who have received immunosuppressive therapy (e.g., recipients of bone marrow or organ transplants) and those with the acquired immunodeficiency syndrome (AIDS) are also at a higher risk for developing NHL. The overall prevalence of lymphomas among children with HIV infection is approximately 1.6%. Among HIV-positive children with hemophilia, NHL is 36 times more frequent than in HIV-negative children with factor 8 deficiency. The majority of the HIV-associated NHLs have a B-cell immunophenotype with either aggressive B-cell/Burkitt or large-cell morphology. Proliferative lesions of mucosa-associated lymphoid tissue (MALT), which may be either benign or malignant, have also been described in children with HIV infection. Although deficient T-cell function has been implicated in these congenital and acquired immunodeficiency states, further study is required to fully clarify the mechanisms of pathogenesis.

There are well-recognized geographic differences in both the incidence and the distribution of histologic subtypes of NHL. For example, NHLs are very rare in Japan but are very common in equatorial Africa. More specifically, Burkitt lymphoma accounts for approximately half of all childhood malignancies in equatorial Africa and is the predominant NHL subtype in northeastern Brazil and areas of the Middle East. In contrast, lymphoblastic lymphomas are the predominant histologic subtype in southern India. In some parts of the world, the distribution of histologic subtypes of childhood NHL has yet to be firmly established.

There are also geographic differences in the clinical and biological features of some NHLs. For example, whereas Burkitt lymphoma in equatorial Africa (endemic Burkitt lymphoma) frequently involves the jaw, abdomen, orbit, paraspinal area, and central nervous system (CNS), the common sites of involvement associated with Burkitt lymphoma in the United States and Western Europe (sporadic Burkitt lymphoma) include the abdomen, bone marrow, and nasopharynx. The predominant chromosome 8 breakpoints, as well as the breakpoints within the immunoglobulin heavy-chain gene (on 14q32) differ between sporadic and endemic cases. In addition, sporadic cases have more complex karyotypic abnormalities than endemic cases in addition to the classic translocations, suggesting distinct mechanisms of malignant transformation. Whereas the endemic cases are associated with IgM expression, the sporadic cases generally are not. The sporadic and endemic cases also differ with respect to Epstein-Barr virus (EBV) association. The overlap of the lymphoma belt in equatorial Africa with the malaria belt prompted speculation that an infectious agent might be involved in lymphomagenesis. This led to the discovery of the EBV and its association with African Burkitt lymphoma. Although a direct role in pathogenesis has not been demonstrated, the circumstantial evidence for its involvement is compelling. It has been suggested that as a B-cell mitogen, EBV increases the target pool of cells potentially susceptible to malignant transformation. Supporting this hypothesis is evidence that Rag gene expression can be induced by EBV, theoretically increasing the likelihood of a translocation occurring in immature B cells that are about to rearrange their immunoglobulin genes. The potential role of Epstein-Barr nuclear antigen (EBNA)-1 in pathogenesis has been suggested by experiments demonstrating that lymphomas develop in mice transgenic for EBNA-1. Moreover, an identified EBNA-1 variant has been shown to be associated with the majority of Burkitt lymphoma cases studied, prompting investigators to speculate that this tumor-associated mutation alters EBNA-1 function in a way that directly or indirectly provides a growth advantage for the lymphoma cell. A more direct role for EBV in lymphomagenesis is suggested by studies of the EBV-positive Burkitt lymphoma cell line, Akata, which loses its malignant phenotype with spontaneous loss of EBV; however, the malignant phenotype is regained with EBV reinfection. EBV association has been reported in approximately 90% of the endemic (i.e., African) Burkitt tumors and in approximately 15% of sporadic cases (United States and Western Europe). Aberrant and disrupted expression of the EBV genome has recently been reported in cases of sporadic Burkitt lymphoma that were EBV-negative by conventional EBNA screening. This observation, coupled with the 50% rate of EBV association in Burkitt tumors in other parts of the world (including Brazil, Russia, Argentina, and Chile), suggest a widespread role for this virus in lymphomagenesis.

Pathology and Biology

The classification systems applied to pediatric NHL are similar to those used for lymphomas occurring in adults. In the National Cancer Institute (NCI) Working Formulation for clinical use, published in 1982, which subclassified NHLs based on their morphologic appearance and clinical aggressiveness into three grades (low, intermediate, and high), most of the childhood NHLs would be designated as high-grade lymphomas ( Fig. 94.1 ). The Revised European-American Lymphoma (REAL) Classification introduced in the mid-1990s and the World Health Organization (WHO) Classification initially published in 2001 and updated in 2008 and 2016 define the different categories of NHL as distinct clinicopathological entities, with well-defined morphologic, immunophenotypic, and genetic features, named, when feasible, according to their postulated counterparts in the normal lymphoid system. This has led to the definition of several lymphoma subtypes within the high-grade lymphoma category previously assigned to most pediatric lymphomas. These include Burkitt lymphoma, diffuse large B-cell lymphoma (DLBCL), precursor B or T lymphoblastic lymphomas, and anaplastic large-cell lymphoma (ALCL). The most recent WHO Classification has added multiple entities as biologically and clinically distinct subtypes of DLBCL, reflecting the significant progress made in better characterizing the lymphomas previously included under this general category (see further below). Some low-grade lymphoma subtypes, including follicular center cell lymphoma and marginal zone lymphoma, may also occur in children where they appear to have distinct clinicopathological features, albeit with less frequency. The WHO classification of pediatric NHLs is summarized in Table 94.1 .