Hodgkin Lymphoma

Monika L. Metzger

Matthew J. Krasin

John K. Choi

Melissa M. Hudson

While the original paper by Hodgkin entitled “On Some Morbid Appearances of the Absorbent Glands and Spleen” already appeared in 1832,¹ it was not until the second half of the 19th century that Sternberg (1898) and Reed (1902) described the histopathology of Hodgkin lymphoma (HL). Only the successful cultivation of Reed-Sternberg cells in the early 20th century permitted the demonstration of the cells’ malignant nature.²

Radiation therapy was the first modality that resulted in the cure of early-stage disease. In 1964, the MOPP regimen (mechlorethamine [nitrogen mustard], Oncovin [vincristine], procarbazine, prednisone), a four-drug non-cross-resistant antineoplastic combination with nonoverlapping toxicities, successfully cured approximately 50% of patients.³ With improved survival after MOPP, investigators appreciated an increased risk of treatment-related acute myeloid leukemia (AML) and infertility, which led to the development of the ABVD regimen (Adriamycin [doxorubicin], bleomycin, vinblastine, dacarbazine) in the 1970s that was not associated with an excess risk of secondary AML or infertility. The lack of leukemogenesis and permanent gonadal toxicity and superior treatment outcomes have led to the standard use of ABVD in adults with newly diagnosed HL. However, the sole use of ABVD in children has been less popular because of concerns about potential cardiopulmonary toxicity.

With greater appreciation of treatment sequelae after standard-dose radiotherapy and non-cross-resistant chemotherapy, pediatric investigators modified treatment strategies in the 1980s to address the specific needs of children. Combined-modality therapy regimens evolved in risk-adapted trials in which patients with favorable clinical presentations received fewer cycles of multiagent chemotherapy and lower radiation doses and treatment volumes. Currently, many pediatric HL study groups use compacted dose-intensive multiagent chemotherapy for patients with advanced and unfavorable disease. Standard incorporation of computed tomography (CT) and 18-fluoro-2-deoxyglucose (FDG) positron emission tomography (PET) scans into early response evaluation are being investigated for therapy reductions or radiation omission in patients who achieve early complete responses (CRs) to initial chemotherapy. Currently, incorporation of targeted therapy into frontline trials is also being investigated.

EPIDEMIOLOGY

HL has a unique bimodal age distribution that differs geographically and ethnically. In industrialized countries, the early peak occurs in the middle-to-late 20s and the second peak after the age of 50 years. In low-income countries, the early peak occurs before adolescence. Epidemiologic studies demonstrate three distinct forms of HL: a childhood form (14 years or younger), a young adult form (15 to 34 years), and an older adult form (55 to 74 years). HL is rarely diagnosed in children younger than 5 years. There is a slight overall male predominance in the childhood form. Among adolescents, the gender distribution is roughly equal. In the United States, the incidence of HL among Whites is slightly higher than that among Blacks, and is the lowest for Asian Americans.⁴

The childhood form of HL tends to increase with increasing family size and decreasing socioeconomic status. In contrast, the young adult form is associated with a higher socioeconomic status in industrialized countries. The risk for young adult HL decreases significantly with increased sibship size and birth order. Swedish investigators observed a lower risk of HL in young adults with multiple older, but not younger siblings, consistent with the hypothesis that early exposure to viral infection may play a role in the pathogenesis of the disease.⁵ Likewise, early exposure to common infections in preschool appears to decrease the risk of HL, most likely by promoting maturation of cellular immunity.⁵ Furthermore, a British case-control study found HL patients to have excess visits to their primary care provider for random infections in the years preceding their diagnosis, possibly reflecting an underlying immune abnormality.⁶ Histologic subtypes also show variability related to age at diagnosis. Mixed cellularity (MC) HL is more common at younger ages, whereas nodular sclerosing (NS) HL has a higher incidence in more affluent societies.

Familial HL

Clustering of cases of HL within families or races may suggest a genetic predisposition to the disease or a common exposure to an etiologic agent. Studies of affected families have suggested an increased association of HL with specific HLA antigens. The concordance of HL in first-degree relatives, especially monozygotic twins, but also siblings, and parent-child pairs has been noted in numerous reports. The elevated risk of HL ranges from threefold among parent-offspring pairs to fivefold among siblings (8-fold among brothers and 11-fold among sisters), and the familial risk being greater in index cases younger than 35 years.⁷^,⁸ Etiologic factors underlying immune deficiency include genetic (e.g., ataxia-telangiectasia), infectious (e.g., human immunodeficiency virus), and iatrogenic agents.⁹^,¹⁰^,¹¹

Epstein-Barr Virus and HL

The epidemiologic characteristics of HL suggest that its etiology may vary by age at presentation and suggests an infectious etiology. A potential causative agent is Epstein-Barr virus (EBV). Enhanced activation of EBV may precede the development of HL because high EBV antibody titers are seen in many patients with HL. This hypothesis is supported by in situ hybridization evidence of EBV genomes and RNA transcripts (EBER1 and EBER2) in Hodgkin Reed-Sternberg (HRS) cells. EBV is detected in HL in variable frequencies, ranging from 3% to 100% depending on histologic subtypes and HIV status.¹²

In EBV-associated HL, the virus is localized to the HRS cell, EBV latent gene products are expressed, and the EBV infection is clonal. HL that is EBV-positive at initial diagnosis is usually also positive at relapse with persistence of the same EBV strain. EBV-infected HRS cells consistently express EBV-encoded latent membrane protein (LMP1), LMP2A, and Epstein-Barr nuclear antigen 1 (EBNA-1), but not EBNA-2, viral capsid antigen, or early antigen. This expression pattern is associated with the latency type II pattern of EBV infection.

The association of EBV and its strain subtypes with HL depends on ethnicity, age, geography, and socioeconomic status of the region. EBV is detected in HRS in 93% of Asian, 86% of Hispanic, 46% of Caucasian, and 17% of African American children with HL. EBV-positive tumor genomes are more frequently observed in children younger than 10 years living in low-income countries. Finally, EBV strain type 1 is predominant in the United Kingdom, South Africa, Australia, and Greece, while EBV type 2 is predominant in Egypt, but both EBV strains can be seen in 21% of cases.

BIOLOGY

HL is a B-lineage lymphoma characterized by a small number of clonal tumor cells (classical HRS cell and its morphologic variants), surrounded by rosettes of T lymphocytes and a polymorphous inflammatory cell population that constitutes the bulk of the tumor tissue. Consequently, the true origin of the neoplastic cells remained uncertain until technical advances to enrich HRS to 5% of total cells permitted detection of clonal IgH rearrangement by Southern blot analysis and confirmed by additional technical advances such as microdissection and single cell polymerase chain reaction (PCR) that permitted independent analysis of HRS cells removed from their benign polyclonal and nonneoplastic inflammatory background. Thus, HL is a unique type of B-cell neoplasm, and hence the current World Health Organization (WHO) classification¹³ has abandoned the term “Hodgkin disease” for “Hodgkin lymphoma” to reflect the true origin of the HRS cells.

Studies focusing on individual HRS cells identified three distinct origins. In one subtype of HL, the nodular lymphocyte-predominant HL (nLPHL), the HRS cells derive from germinal center (GC) or post-GC B cells and show rearranged immunoglobulin variable (IgV) genes. The presence of intraclonal IgV gene diversity further indicates an origin in mutated and antigen-selected GC B cells that have retained expression of all B-cell-specific molecules, such as CD19, CD20, CD79a, J chain, PAX5, Ig (with light-chain restriction), OCT2, BOB1, and PU-1, although at reduced levels of expression.¹⁴^,¹⁵ In the other subtypes of HL, the HRS cells are also GC B cells, but they have crippling DNA mutations that destroy the coding capacity of their previously functional IgV gene rearrangements. Expressions of other B-cell-specific genes are greatly diminished or lost¹⁶ presumably by other mutations or epigenetic processes. These mutations and decreased expression will target normal GC B cells for apoptosis, an important cell regulatory process that must be subverted in HL.

Antiapoptotic Mechanisms in HRS

The inhibition of apoptosis in HRS cells has been attributed to a variety of genetic and signaling alterations,¹⁷^,¹⁸^,¹⁹ including activation of NOTCH-1, STAT (STAT 3, 5, and 6), multiple receptor tyrosine kinases, AP-1 transcription factor, and NFκB signaling pathway. NFκB activation is mediated by EBV in HRS cells and mediated in EBV-negative HRS by inactivating mutations of the IκB family members that are negative regulators of NFκB.²⁰

NFκB, AP-1, and other cell signaling molecules are altered by the CD30 molecule that is highly expressed by HRS cells. The CD30 antigen was initially recognized on the HRS cells using the Ki-1 monoclonal antibody. The CD30 molecule is a membrane glycoprotein that belongs to the tumor necrosis factor receptor superfamily. The extracellular domain of this receptor molecule binds the CD30 ligand, while its intracellular portion signals through the tumor necrosis factor (TNF) receptor-associated factor pathways, leading to activation of the NFκB transcription factor.²¹ While CD30 expression is seen in almost all HRS cells, it is not specific and is expressed on a variety of normal reactive lymphoid cells (both B and T) and in a variety of B- and T-cell non-HL subtypes.

While the exact roles and regulations of CD30 in classical HL development are still being elucidated, the serum levels of soluble CD30 have been found to correlate with outcome in advanced stage HL. Elevated serum levels of CD30 have been associated with advanced stage, constitutional (B) symptoms, and poor outcome.²² In addition, CD30 is a suitable molecular target in the immunotherapy of HL and other CD30-positive lymphomas.²³^,²⁴

In a subset of HL cases, EBV likely plays an important role in the pathogenesis and development of HL. EBV-infected HRS cells express high levels of LMP1, a viral protein that resembles a constitutively activated member of the TNF receptor superfamily. LMP1 activates a variety of signaling apoptotic and growth pathways, including the transcription factor NFκB.²⁵ LMP1 increases the expression of bcl2, interleukin-10 (IL-10), and major histocompatibility complex class I proteins. EBV-positive cases of HL have higher levels of IL-10, EBV-specific cytotoxic T cells, and MHC class I molecules than EBV-negative cases.²⁶ EBV also increased expression of LMP2A, another EBV-related protein, in HRS cells, and interferes with normal B-cell development by blocking B-cell receptor (BCR)-signaling.²⁵ Finally, EBV infection can rescue normal GC B cells that are destined for apoptosis, suggesting that EBV is sufficient to rescue HRS from apoptosis.²⁷

Cytokine Secretion Leading to Prominent Inflammatory Background

Activation of the NFκB and c-Jun N-terminal kinase pathways in HRS cells leads to secretion of many cytokines that recruit inflammatory cells.²¹ Distinct patterns of cytokine production are associated with different subtypes of HL, resulting in a distinct histologic appearance and spectrum of clinical features in each of these types. The histopathologic features of HL with eosinophilia and collagen sclerosis are attributed to HRS cells that secrete cytokines such as IL-4, IL-5, CCL11 (eotaxin), IL-6, IL-7, IL-13, TNF, lymphotoxin, transforming growth factor-β (TGF-β), and basic fibroblast growth factor. These cytokines also induce adhesion molecules that influence the interaction of HRS cells with the neighboring T lymphocytes and their metastatic capacity. Serum levels of IL-6 and TGF-β correlate well with systemic symptoms and immunosuppression, respectively. Table 22.1 summarizes the relationship among cytokine production and common clinical and pathologic features of HL.

Complexities and Future Possibilities

While most of the discussion of the biology of HL has focused on HRS cells and apoptosis, the whole story is likely to be much more complicated. In addition to the HRS cells, nonneoplastic inflammatory cells also contribute to the disease. This microenvironment has prognostic significance as increased numbers of CD68- or CD163-positive macrophages correlate with poor prognosis in some adult studies.¹⁹^,²⁸^,²⁹^,³⁰ The background T cells, fibroblasts, macrophages, and neutrophils produce cytokines CCL5, CCL11, and CXCL8(IL8), respectively.³¹ In addition to alteration in apoptosis, differentiation, and cytokine production, other cell functions are additionally affected.¹⁹

Progress in HL understanding has been hampered by the scarcity of HRS in the tumor, requiring mostly single HRS cell analysis by immunohistochemistry or arduous microdissection followed by PCR analysis. Recently, large numbers of HRS cells could be isolated by magnetic or flow cytometry-based sorting of dissociated HL samples using one to multiple antibodies.³²^,³³^,³⁴ This approach should permit next generation sequencing of the HRS cells at very high resolution. As these sequencing endeavors add to the ever-growing potential driver mutations and therapeutic targets in HL, better functional experimental models need to be developed. While some HRS cells lines are available for in vitro preclinical
studies, animal models that reliably recapitulate human HRS³⁵ or xenograft models are still lacking.

TABLE 22.1 Clinical and Pathologic Features of Hodgkin Lymphoma Related to Cytokine Production

Clinical and Pathologic Features of Hodgkin Lymphoma	Cytokines
Constitutional (“B”) symptoms	TNF, LT-α, IL-1, IL-6
Polykaryon formation	Interferon-γ, IL-4
Sclerosis	TGF-β, LIF, PDGF, IL-1, TNF
Acute-phase reactions	IL-1, IL-6, IL-11, LIF
Eosinophilia	IL-5, granulocyte M-CSF, IL-2, IL-3
Plasmacytosis	IL-6, IL-11
Mild thrombocytosis	IL-6, IL-11, LIF
T-cell and Hodgkin and Reed-Sternberg cell interaction	IL-1, IL-2, IL-6, IL-7, IL-9, TNF, LT-α, CD30L, CD40L, B7 ligands (CD80 and CD86)
Immune deficiency	TGF-β, IL-10
Autocrine growth factors	IL-6, IL-9, TNF, LT-α, CD30L, M-CSF
Increased alkaline phosphatase	M-CSF
Neutrophil accumulation/activation	IL-8, TNF, TGF-β
IL, interleukin; LIF, leukemia inhibitory factor; LT, lymphotoxin; M-CSF, macrophage colony-stimulating factor; PDGF, platelet-derived growth factor; TGF, transforming growth factor; TNF, tumor necrosis factor.
Adapted from Kadin ME, Liebowitz DN. Cytokines and cytokine receptors in Hodgkin’s disease. In: Mauch PM, Armitage JO, Diehl V, et al, eds. Hodgkin’s disease. Philadelphia, PA: Lippincott Williams & Wilkins, 1999:139.

PATHOLOGY

HL is characterized by a few scattered neoplastic HRS cells that account for 0.1% to 10% of the total cell population of the tumor infiltrate. Most of the infiltrate is composed of nonneoplastic background inflammatory cells (histiocytes, plasma cells, lymphocytes, eosinophils, neutrophils) and fibrosis that develop as a result of cytokines. HL must be differentiated from several subtypes of NHL with similar morphologic characteristics (T-cell-rich large B-cell lymphoma and anaplastic large cell lymphoma) and from benign lymphoid hyperplasia with a similar cellular composition (e.g., infectious mononucleosis, progressive transformation of GCs).³⁶ In addition, less common variants of HL (e.g., lymphocyte depleted and syncytial variants) may overlap morphologically with nonhematopoietic malignancies, such as carcinomas and sarcomas. With the advent of widely available immunohistochemical staining, the latter entities are usually easy to exclude.

Definition of Histologic Subtypes

The current WHO histologic classification of HL is largely based on the Rye modification of the Lukes and Butler classification schema. The WHO classification¹³ recognizes two major subtypes of HL—classical HL (cHL) (Figs. 22.1a and 22.2a) and nodular lymphocyte-predominant HL (nLPHL) (Figs. 22.1b and 22.2b)—based on their histology and immunohistochemical staining pattern (Tables 22.2 and 22.3). For example, the HRS cells in nLPHL are usually mononuclear with markedly convoluted and lobated nucleus (“popcorn cells”), thin nuclear membrane, pale chromatin, and one-to-several small basophilic nucleoli (Fig. 22.1b). These HRS variants are also called lymphocytic and histiocytic (L&H) cells. In contrast, the HRS cells of cHL are typically large (≥15 to 45 µm in diameter), with abundant slightly basophilic cytoplasm, two nuclei or two nuclear lobes, and prominent eosinophilic large nucleoli (Fig. 22.1a).

Figure 22.1 Morphologic features of the neoplastic cells in Hodgkin lymphoma (HL). A: Reed-Sternberg cell in classical HL. B: Reed-Sternberg cell in nodular lymphocyte predominant HL. (Hematoxylin & eosin stain; original magnification 600x, oil immersion).

Figure 22.2 Morphologic patterns of lymph node infiltrate in HL. A: Cellular nodules are surrounded by dense fibrous bands in classical HL, nodular sclerosis subtype. B: Cellular nodules without fibrosis in nodular lymphocyte predominant HL (Hematoxylin & eosin; original magnification 20×, dry).

cHL is further subdivided into four histologic subtypes based on the tissue architecture, the presence of fibrosis, and the features of the associated inflammatory infiltrate (Table 22.2). Historically, prognosis for some of the categories of HL was linked to the ratio of lymphocytes to abnormal cells. Since the development of highly curative treatment regimens, however, all histologic subtypes of HL are equally responsive to treatment.

In all four subtypes of cHL, the HRS cells are B cells with clonal immunoglobulin heavy-chain rearrangement but lack many immunohistochemical evidence of B-lineage differentiation. The HRS cells are negative for leukocyte common antigen (CD45), B-lineage antigens (such as CD19, CD20, CD79a, OCT2, BOB1, and J chain), and BCL6. In a limited number of cases (10%-20%), the cHRS cells may express CD20 or CD79a with variable intensity. Of the B-cell-specific antigens detected by immunohistochemistry, PAX5/BSAP is the most useful as 98% of cHL cases have dim expression of PAX5 in the HRS cells. HRS cells retain molecules important in B-T-cell interactions, such as CD80, CD86, and MHC class II, and often express molecules typically upregulated in plasma cells (CD138 and MUM-1).³⁷^,³⁸ HRS cells characteristically coexpress the non-B-cell-specific CD15 (LeuM1) and CD30 (Ki-1) antigens. Finally, in a minority (2%) of cHL cases, the HRS cells show a cytotoxic T-lymphoid immunophenotype that may be associated with a B-cell genotype or harbor T-cell receptor beta gene rearrangements.³⁷^,³⁹^,⁴⁰

TABLE 22.2 Histological Classification of Hodgkin Lymphoma (HL) According to the WHO Classification

	Prevalence in Children %	Reference
Nodular lymphocyte predominant HL	<5	148
Classical HL
Nodular sclerosis subtype	40-70
Mixed cellularity subtype	30	149
Lymphocyte-rich subtype	<5
Lymphocyte depleted subtype	<5

TABLE 22.3 Immunophenotype of the Neoplastic Cells and Pathologic Features in Hodgkin and non-Hodgkin Lymphomas

Tumor Type	LCA (CD45)	PAX5	CD30 (Ki-1)	CD20	CD3	ALK	Growth Pattern	Inflammatory Cells
cHL, HRS cell	−	+ dim	+	−/+	−	−	Nodular, diffuse	Mixed
nLPHL, L&H cell	+	+ dim	−	+	−	−	Nodular +/− diffuse	Small B & T cell
T−cell rich large B−cell lymphoma	+	+	−	+	−	−	Diffuse	Small T cells
ALCL, T/null cell	+	−	+	−	+/−	+/−	Diffuse, sinusoidal, interfollicular	Mixed
cHL, classical Hodgkin lymphoma; HRS, Hodgkin Reed-Sternberg; nLPHL, nodular lymphocyte-predominant Hodgkin lymphoma; ALCL, anaplastic large cell lymphoma.

Unlike cHL, nLPHL have HRS cells that consistently show immunophenotypic evidence of B-lineage differentiation and express CD45, CD20, CD79a, J chain, OCT2, and BOB1, in addition to PAX5. Furthermore, HRS cells of nLPHL lack expression of non-B-cell antigens CD15 and CD30, typically seen in cHRS.

Complexities

Gray-Zone Lymphoma

The diagnoses of cHL and nLPHL are straightforward in the vast majority of cases and determined by histology and immunophenotyping. However, rare cases are challenging in distinguishing HL from NHL of B-cell lineage because the morphologic and immunophenotypic features overlap with those of HL and diffuse large B-cell lymphoma. These are often designated “gray-zone lymphomas”⁴¹ and generate anxiety and robust discussion among pathologists.

nLPHL versus cHL

HRS cells in nLPHL have different morphology and immunophenotype compared to those in cHL. Furthermore, nLPHL has a
propensity to transform to NHL, usually diffuse large B-cell lymphoma, in 10% to 14% of cases.⁴² Despite these dissimilarities, nLPHL has similar contiguous lymph node spread and favorable response to typical HL chemotherapy to those seen with cHL. Case reports suggest that these two entities can share more similarities than anticipated. For example, HRS cells, in many cases of nLPHL, have decreased/absent expression of the B-cell antigens CD19, CD22, Lyn, Ets-1, LCK.¹⁵ Even EBV-infected HRS initially thought to be lacking in nLPHL have now been reported in occasional cases of nLPHD.¹² Finally, composite lymphomas consisting of nLPHL and cHL, as defined by the typical morphologies and immunophenotypes of the two subentities, have been reported in a single patient⁴³ or even in a single lymph node.⁴⁴ In these composite lymphomas, the HRS cells of the two subtypes of HL were microdissected and had identical IgH rearrangements using PCR amplification and multiple primer sets including sequence specific primers.⁴⁴

CLINICAL PRESENTATION

Lymphadenopathy

Most patients present with painless cervical or supraclavicular adenopathy. Affected lymph nodes are firmer than inflammatory nodes, feel rubbery, and rarely may be sensitive to palpation if they have grown rapidly. At least two-thirds of patients present with some degree of mediastinal involvement, which may cause a nonproductive cough, orthopnea, or other symptoms of tracheal or bronchial compression. Posteroanterior and lateral thoracic radiographs should be performed as soon as HL becomes part of the differential diagnosis (presence of a supraclavicular mass) and to assess airway patency. In younger children, mediastinal lymphadenopathy may be difficult to distinguish from a large, normal thymus. Infrequently, axillary or inguinal lymphadenopathy is the first presenting sign. Primary disease presenting in a subdiaphragmatic site is rare and occurs in only approximately 3% of cases.

Systemic Symptoms

Nonspecific systemic symptoms may include fatigue, anorexia, and slight weight loss. Three specific constitutional (“B”) symptoms correlate with prognosis: unexplained fever with temperatures above 38°C orally, unintentional weight loss of more than 10% within 6 months preceding diagnosis, and drenching night sweats. Pruritus can sometimes also be observed and may be mild or severe enough that scratching can cause excoriations, particularly in women with advanced disease. Another unusual and so far unexplained syndrome associated with HL is alcohol-induced pain in areas of nodal enlargement that begins within minutes of drinking alcohol.

Laboratory Profile

Hematologic and chemical blood parameters show nonspecific changes that may correlate with disease extent. Abnormalities of peripheral blood counts may include neutrophilic leukocytosis, lymphopenia, eosinophilia, and monocytosis. At the onset of disease, the absolute lymphocyte count is usually normal in children, although adults with extensive disease commonly have lymphopenia. Anemia may indicate the presence of advanced disease, and usually results from impaired mobilization of iron stores. Hemolytic anemia associated with HL may be Coombs positive and is accompanied by a reticulocytosis and normoblastic hyperplasia of the bone marrow.

Several autoimmune disorders have been observed in association with HL, including nephrotic syndrome, autoimmune hemolytic anemia, autoimmune neutropenia, and immune thrombocytopenia (ITP). ITP has been reported in 1% to 2% of cases of HL and may occur in association with autoimmune hemolytic anemia. Thrombocytopenia may develop before, at the same time, or after the diagnosis of HL. ITP frequently occurs in patients in remission after completion of therapy for HL and is not usually associated with relapse. The treatment approach recommended for ITP in patients with HL is similar to that in patients without malignancy. Response to ITP therapy is also similar.

The erythrocyte sedimentation rate, serum copper, and ferritin levels may be elevated, reflecting activation of the reticuloendothelial system. C-reactive protein is another acute-phase reactant produced in the liver.⁴⁵ These nonspecific tests, if abnormal at diagnosis, may be useful in follow-up evaluation.

Immunologic Status

Patients with HL exhibit a variety of immune system abnormalities at diagnosis that may persist during and after therapy. Natural killer cell cytotoxicity may be reduced in untreated patients. Typically, enhanced sensitivity to suppressor T lymphocytes present at diagnosis, results in abnormal cellular immunity. After treatment, humoral immunity may be transiently depressed. In vitro studies have provided insights regarding the mechanism of immune dysregulation in HL. Several cytokine interactions have been proposed to explain the paradoxical presence of extensive inflammatory infiltrate, ineffective host antitumor response, and generalized cellular immune deficiency.

DIFFERENTIAL DIAGNOSIS

HL must be differentiated from other malignant lymphomas that can have similar presentations (lymphoblastic lymphoma and diffuse large B-cell lymphomas presenting with a large anterior mediastinal mass) or similar histologic appearance (CD30-positive anaplastic large cell lymphomas). In general, however, the growth rate of these affected lymph nodes is often more rapid than in HL and are accompanied by elevated uric acid or lactic dehydrogenase (LDH) levels (see Chapter 24). Furthermore, soft tissue sarcomas and germ cell tumors can present in the neck or mediastinum, as can metastatic adenopathy from other primary tumors (e.g., nasopharyngeal carcinoma and soft tissue sarcoma); infectious causes of lymphadenopathy, particularly those with an indolent course (e.g., EBV, atypical mycobacterium, histoplasmosis, and toxoplasmosis) need to be considered. Another benign entity that can sometimes be difficult to differentiate from HL is progressive transformation of GCs.⁴⁶ A more complex problem is that of a mediastinal mass that must be differentiated from normal thymus in an otherwise asymptomatic patient. Ultimately, only biopsy can definitively confirm a diagnosis.

DIAGNOSTIC WORKUP

Table 22.4 shows the recommended steps in the diagnostic workup of a child with HL. An excisional lymph node biopsy is the preferred procedure to establish the diagnosis, as it permits evaluation of the malignant HRS cells within the background of characteristic architectural changes associated with the specific histologic subtypes. A careful physical examination with assessment of all node-bearing areas, including Waldeyer ring, is essential, with measurement of enlarged nodes so changes can later be quantified.

Evaluation by an ear, nose, and throat specialist can sometimes be helpful, and a contrast-enhanced CT of the neck is always recommended. The chest radiograph provides preliminary information about mediastinal involvement and intrathoracic structures. Patients are considered to have “bulky” mediastinal lymphadenopathy if it measures greater than or equal to 33% of the maximum intrathoracic cavity. The pulmonary parenchyma, chest wall, pleura, and pericardium are the most commonly involved extranodal sites of disease and should be further assessed by CT.

TABLE 22.4 Diagnostic Evaluation of Children with Hodgkin Lymphoma

Diagnostic Evaluation	Important Elements	Comments
Medical history	“B” symptoms	Unexplained fever with temperatures above 38°C orally, unexplained weight loss ≥10% within 6 months preceding diagnosis, and drenching night sweats
	Symptoms of a large mediastinal mass	Superior vena cava syndrome: dyspnea, facial swelling, cough, orthopnea, and headache Tracheal or bronchial compression: cough, dyspnea, and orthopnea
Physical examination	Lymph nodes	Location and size
	Tonsils	Symmetry, size, and nodularity
	Lung auscultation	Stridor and wheezing
	Abdomen	Hepatomegaly and splenomegaly
Laboratory tests	Complete blood count	Anemia, leukocytosis, and lymphopenia
	Biochemistry profile	Elevated lactate dehydrogenase and low albumin. Renal and hepatic function studies prior to starting chemotherapy to determine need for dose adjustments and tolerability. Elevated alkaline phosphatase is associated with bone involvement.
	Erythrocyte sedimentation rate, C-reactive protein, serum ferritin	Elevation of markers of inflammation at diagnosis can be followed for response to therapy.
Diagnostic imaging, anatomic	Chest radiography	Mediastinal-to-thoracic ratio
	CT of neck, chest	Location and size of lymph nodes to evaluate pulmonary involvement.
	CT or MRI of abdomen and pelvis	Location and size of lymph nodes and hepatosplenic involvement.
Diagnostic imaging, functional	FDG-PET	Metabolic activity of involved nodes and extralymphatic organs, including bone and bone marrow. High sensitivity and low specificity.
Biopsy	Lymph node Bone marrow	Histologic confirmation of Hodgkin lymphoma Mostly being replaced by FDG-PET

The presence of infradiaphragmatic disease is most frequently evaluated by CT with both oral and intravenous contrast agents to accurately delineate lymphadenopathy from other infradiaphragmatic structures. Evaluation of the extent of abdominal and pelvic disease by CT scan is further complicated in children by the lack of retroperitoneal fat.

Splenic involvement occurs in 30% to 40% of patients with HL, and the size of the spleen may not correlate with the degree of disease involvement. Liver size and liver function studies are also unreliable indicators of hepatic disease. Both CT and MRI scans may suggest splenic or hepatic involvement when these organs appear enlarged with areas of abnormal density.⁴⁷ Occasional characterization of the visualized lesions by ultrasound may also be useful.

Functional nuclear imaging studies are routinely used as a diagnostic and monitoring modality. PET is the preferred functional imaging modality for lymphoma staging.⁴⁸^,⁴⁹^,⁵⁰^,⁵¹^,⁵²^,⁵³^,⁵⁴ The integration of functional and anatomic tumor characteristics provided by PET-CT imaging has popularized its use for staging and monitoring of pediatric patients with lymphoma because it is both an accurate and cost-effective modality. In PET scanning, uptake of the radioactive glucose analogue, FDG correlates with proliferative activity in tumors undergoing anaerobic glycolysis. PET-CT combinations can be very helpful in determining whether residual mass-like opacities on CT represent active disease or areas of fibrosis.⁵⁵ Residual or persistent FDG avidity appears to be useful in predicting prognosis and the need for additional therapy in post-treatment evaluation.⁴⁹^,⁵⁴ However, FDG-PET also has limitations in the pediatric setting.⁵¹ Tracer avidity may be seen in a variety of nonmalignant conditions, including thymic rebound commonly observed after completion of lymphoma therapy. FDG avidity in normal tissues, for example, brown fat of cervical musculature, may confound interpretation of the presence of nodal involvement by lymphoma. Lastly, tumor activity cannot be correlated with FDG in patients with diabetes who do not have well-controlled blood glucose levels.

Primary bone involvement in HL is rare; however, it may occur in 5% to 20% of patients during the course of disease.⁵⁶ It is confirmed in a child who has bone pain, elevated serum alkaline phosphatase concentration beyond that expected for age, or extranodal (bony) disease identified by FDG PET. FDG-PET scans are much more sensitive and specific then technetium-99 bone scans and have therefore completely replaced them.

Bone marrow involvement at the time of initial presentation of HL is uncommon and rarely occurs as an isolated site of extranodal disease. The pattern of infiltration in the bone marrow may be diffuse or focal and is often accompanied by reversible marrow fibrosis. Bone marrow biopsies have typically been performed in patients with clinical stage III to IV disease or B symptoms, or in any patient at the time of disease recurrence. However, there is evidence that FDG PET may be more sensitive and specific for the detection of bone marrow disease than a biopsy, given the focal pattern of involvement that may be missed by an iliac crest biopsy and may safely substitute it altogether.⁵⁷

STAGING

HL appears to spread along contiguous lymph nodes until late in the course of disease. The currently used Ann Arbor staging system, adopted in 1971, is based on this observation (Table 22.5). The anatomic locations of lymph node chains designated as regions for the purpose of staging are illustrated in Figure 22.3. The substage classifications A, B, E, S, and X amend each stage based on defined clinical features. Substage A indicates “asymptomatic” disease. B symptoms include fever exceeding 38°C for three consecutive days, drenching night sweats, and an unexplained weight loss of at
least 10% over 6 months. Substage E denotes extralymphatic extension from contiguous nodal disease. Disease involving the spleen is designated an S, and mediastinal bulk is often designated with an X.

TABLE 22.5 Ann Arbor Staging Classification for Hodgkin Lymphoma

Stage	Definition
I	Involvement of a single lymph node region (I) or of a single extralymphatic organ or site (I_E)
II	Involvement of two or more lymph node regions on the same side of the diaphragm (II), or localized involvement of an extralymphatic organ or site and one or more lymph node regions on the same side of the diaphragm (II_E)
III	Involvement of lymph node regions on both sides of the diaphragm (III), which may be accompanied by involvement of the spleen (III_S) or by localized involvement of an extralymphatic organ or site (III_E) or both (III_SE)
IV	Diffuse or disseminated involvement of one or more extralymphatic organs or tissues with or without associated lymph node involvement
The absence or presence of fever higher than 38°C for three consecutive days, drenching night sweats, or unexplained loss of 10% or more of body weight in the 6 months preceding admission are to be denoted in all cases by the suffix letters A or B, respectively. Suffix letter E denotes extralymphatic extension from contiguous nodal disease. Disease involving the spleen is designated a suffix letter S, and mediastinal bulk is often designated with an X.

Figure 22.3 Anatomic definition of separate lymph node regions used for staging purposes. (Adapted from Kaplan HS, Rosenberg SA. The treatment of Hodgkin’s disease. Med Clin North Am 1966;50:1591.)

TREATMENT

Principles of Radiotherapy

Incorporation of radiation therapy in the management of pediatric HL has evolved over the past 30 years from single modality approaches incorporating high-dose extensive nodal irradiation or aggressive multiagent systemic chemotherapy to combined-modality therapy. Multiple clinical trials have addressed the additive value of involved field radiation to multiagent chemotherapy, suggesting a benefit for the population of HL patients as a whole.⁵⁸^,⁵⁹^,⁶⁰ Based in part on these studies, a risk-adapted approach has been developed that attempts to tailor the amount and intensity of therapy, both chemotherapy and particularly radiation, to the severity of disease.⁵⁸^,⁶¹^,⁶² Several important questions remain in the integration of radiation for HL occurring in children and adolescents. Selection of patients for radiation avoidance, integration of modern radiotherapeutic approaches for improved conformal delivery of dose, and further reduction of the volume of treatment are current questions in modern day clinical trials that attempt to maintain the excellent disease control rates already achieved with traditional radiation therapy approaches, while further minimizing the long-term risks associated with wide field radiation exposures.⁶³ Understanding the principles and approaches of radiation therapy employed in existing and published studies as well as the opportunities for improvement in the delivery of this modality will be important for pediatric oncologists and radiation oncologists as we continue to make progress in this highly curable disease.

Selection of Dose and Timing

Dose Selection

Selection of radiation dose tends to be trial and group-specific, although the dose ranges are similar, as noted above. Clinical trials have employed 20, 21, and 25.5 Gy,⁵⁹^,⁶⁰^,⁶² using daily doses of 2, 1.75, and 1.5 Gy, respectively. Local control rates remain high, 89% to 95%, despite the lower radiation dose and reduced intensity of systemic therapy.⁶⁰^,⁶⁴ Modifications to these doses are made based on the response to systemic therapy, employing specific time points and methods for response assessment. Dose elimination, dose reduction, and dose escalation are all used in modern clinical trials, but the approach from one trial is not transferable to another. It is important to understand the methods and timing of response assessments as one compares each clinical trial or treats with its specific approach (Table 22.6). Despite the differences in radiation therapy across the studies, several prognostic factors emerge. Bulky mediastinal disease predicts a higher risk of local failure, 21.6% versus 6%, at 5 years.⁶⁴ Higher stage patients experienced greater failure rates, particularly in the absence of radiation therapy, despite favorable responses.⁵⁸ On the basis of this greater understanding of response to risk-adapted therapy, current clinical trials in all studies are looking to better select patients for radiation avoidance, maintaining high nodal control rates while continuing to reduce therapy.

Volume Selection

Significantly less investigated is the role of volume reduction in radiation therapy treatment fields. The “standard” involved field grew out of the components of the classic mantle field, used for treatment of HL with radiation therapy alone. In the combined modality era, treatment volumes no longer need to account for the potential spread of all microscopic disease, focusing more on areas of large volume disease or poor response. The definition of a standard involved field is not standard at all, but attempts have been
made to standardize it.⁶⁵ Definitions of involved fields in pediatric trials for HL have varied and must be understood by the radiation oncologist in the context of the specific chemotherapy being delivered. Figure 22.4 shows a typical involved field defined in the earlier HL collaborative trials.⁶² The German Pediatric Hodgkin Study Group (GPOH-HD) was the first to deliver less than standard involved fields in a clinical trial for children with HL.⁶⁶ Radiation planning was developed in conjunction with a centralized review of imaging resulting in a standardized radiotherapeutic approach across many institutions and disease presentations. The importance in moving beyond standard involved fields is not to improve the local control rates (which already exceed 90%), but to further reduce late effects from radiation therapy. The simplest method for reduction in end-organ toxicity such as hypothyroidism or secondary breast cancer is to avoid treating the organ at risk. The implementation of more conformal fields is progress toward this goal, treating only the individual lymph nodes with a margin for microscopic disease (Fig. 22.5). This in conjunction with modern imaging will continue to reduce exposure of normal tissue to radiation, while maintaining equivalent local disease control rates.⁶⁷ Recent clinical trials have tested a more conformal approach including limited volume fields as depicted in Figure 22.5, and ongoing trials are employing true conformal therapy techniques in this disease (Fig. 22.6).

TABLE 22.6 Radiation Dose Modifications in Modern Trials for Pediatric HL

Study/Group	Standard Dose/Daily Fraction Size	Response Time	Imaging Response Method	Modified Dose
Hodgkin collaborators	25.5/1.5 Gy	8 wk	CT	If CR:
			FDG-PET		No radiation all stages of patients
			Product of perpendicular diameter	If <CR:
			CR = >75% reduction		25.5 Gy—only to the nonresponding LN with a clinical margin with conformal technique
German pediatric Hodgkin studies	19.8/1.8 Gy	8 wk	CT	If CR:
			US		No radiotherapy
			MR	If <CR:
			3D measurement		19.8 Gy to the nodal region
			CR = 95% volumetric reduction
Children’s Oncology Group	21/1.5 Gy	6 wk	CT	If CR/VGPR:
			Product of perpendicular diameter		randomize 0 or 21 Gy
			CR = >80% reduction		If PR:
			VGPR = >60%		21 Gy
CR, complete response; PR, partial response; VGPR, very good partial response.

Figure 22.4 Classic involved-field treatment for a patient with clinical stage IB Hodgkin lymphoma.

Radiotherapy Techniques

Radiation simulation is the process of positioning the patient in a consistent, comfortable and reproducible position for daily therapy. Simulation is accomplished on a CT-based simulator, often with intravenous contrast administered to allow definition of the vasculature next to which the lymphatics reside. CT simulation allows a three-dimensional or volumetric rendering of the patient for mapping of the regions to be radiated as well as adjacent normal tissues.
CT simulation and subsequent volumetric treatment planning require dose calculation for the imaged patient volume and can yield dose information for any point or volume in the patient’s image (Fig. 22.6). Two advantages come from this technique: (1) Doses can be homogenized to a greater degree across the patient’s treated volume compared to prior two-dimensional techniques, resulting in less dose variation from the prescribed dose and potentially less toxicity, and (2) dose can be assessed for any target or normal tissue that is delineated (contoured) on CT. This improved knowledge of specific doses to an organ may help better understand end-organ toxicities that result from radiation as well as allowing for avoidance when possible. When the chest or abdomen is part of the treatment volume it may be appropriate to obtain a CT simulation to assess motion of the mediastinum, hila, diaphragm, and spleen. This study, called a 4D-CT (describing the 4th dimension of time during which targets and organs move), is obtained at the same time as the CT simulation. Targeting for treatment occurs in a “virtual” or simulated computer environment. Nodal regions and other tissues that need treatment as well as normal tissues for avoidance or dose calculation are mapped on each CT image slice. These targets and normal tissues (referred to as contours) are then rendered as three-dimensional shapes. Treatment beams are then created with the aid of the medical physicist in the computer environment to treat the areas at risk and avoid adjacent normal tissues. Beam arrangements are often from anterior and posterior but are not limited to these orientations; current clinical trials are investigating the safety and benefit of more complex beam arrangements. Prior to treatment of disease in the pelvic region it may be necessary to transpose the ovaries medially or use a testicular shield to maintain hormone production and fertility.

Only gold members can continue reading. Log In or Register to continue