Chapter 97: Hodgkin Lymphoma

Classical Hodgkin lymphoma (cHL) is a neoplasm of lymphoid tissue, in most cases derived from a germinal center B cell, defined by the presence of the malignant Hodgkin and Reed-Sternberg cells with a characteristic immunophenotype and appropriate cellular background. cHL accounts for 95 percent of cases and contains four histologic subtypes distinguished on the basis of microscopic appearance and relative proportions of Hodgkin and Reed-Sternberg cells, lymphocytes, and fibrosis (nodular sclerosis, mixed cellularity, lymphocyte-rich and lymphocyte-depleted; Table 97–1). Nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL) represents the other major category, which is distinguished by Hodgkin and Reed-Sternberg variants termed lymphocytic and histiocytic cells that, unlike cHL, express typical B-lineage markers.

In his historic 1832 paper entitled On Some Morbid Appearances of the Absorbent Glands and Spleen, Thomas Hodgkin described the clinical histories and gross postmortem findings of seven cases of the disease that was later to bear his name.¹ In 1856, Samuel Wilks independently described 10 cases of “a peculiar enlargement of the lymphatic glands frequently associated with disease of the spleen,” including four of Hodgkin's original cases.² Upon discovering Hodgkin's original report, he used the appellation “Hodgkin's Disease” in a subsequent series of 15 cases published in 1865.² Thirteen years after Hodgkin's original paper, the first cases of leukemia were described. Cases in which the neoplastic cells remained confined to the lymphatic system were described by Dreschfield (1892)³ and Kundrat (1893)⁴; Kundrat gave the name lymphosarcoma to these cases. The description of additional members of the lymphoma–leukemia complex continued up to the present time.

Carl Sternberg (1898)⁵ and Dorothy Reed (1902)⁶ are credited with the first definitive and thorough descriptions of the pathology of cHL, although a number of investigators from England, Germany, and France had previously recognized the characteristic multinucleated giant cells. In 1926, Fox examined microscopic sections from the gross specimens preserved in the Gordon Museum of Guy's Hospital in London of three of Hodgkin's original cases.⁷ It is remarkable that the preserved microanatomy allowed him to confirm the histopathologic diagnosis in two of these cases. Jackson and Parker made the first serious effort at the histopathologic classification of cHL, correlating their findings with prognosis.⁸ A second advance was made in 1966 when Lukes, Butler, and Hicks proposed a classification that related well to clinical presentation and course.⁹ Their proposal was slightly modified into the Rye classification, in which four histopathologic subtypes were described: lymphocyte-predominant, nodular sclerosis, mixed cellularity, and lymphocyte-depleted. In the World Health Organization Classification of Lymphoid Neoplasms, the NLPHL subtype is clearly distinguished from cHL.¹⁰ The “lymphocyte-rich” subtype of cHL was introduced in 1999.

Peters described a clinical staging system in 1950, emphasizing the diagnostic evaluation of the anatomic extent of disease.¹¹ In 1952, Kinmouth introduced lower-extremity lymphangiography that allowed roentgenologic visualization of the pelvic and retroperitoneal lymph nodes and was found to be far more sensitive than palpation or other radiographic methods.¹² The frequency of unsuspected splenic involvement was revealed in a group of 65 patients subjected to laparotomy and splenectomy with biopsy of splenic hilar, paraaortic and mesenteric nodes, and liver at Stanford University.¹³ These diagnostic procedures led to improved understanding of the mode of dissemination of the disease and correlated well with prognosis, culminating in the modern concepts of staging codified at the Rye, New York, conference in 1965,¹⁴ and further refined at the Workshop on the Staging of Hodgkin's Disease in Ann Arbor, Michigan, in 1971.¹⁵

Pusey (1902)¹⁶ and Senn (1903)¹⁷ were the first to report dramatic regressions of lymphadenopathy with exposure to X-rays, discovered by Roentgen in 1896. Based upon the nearly inevitable recurrence in untreated areas, Gilbert proposed the systematic treatment of both involved and uninvolved areas in 1939.¹⁸ Peters (1950) is credited for the first demonstration of the curative potential of radiotherapy in her classic paper.¹¹ The development of megavoltage radiotherapy (doses >4000 cGy), as reported by Kaplan in 1962,¹⁹ permitted the delivery of tumoricidal doses to virtually all lymphoid regions in the body within acceptable limits of normal tissue tolerance.

The chemotherapy of cHL originated as a byproduct of the wartime work on the mustard gases.^20,21 Following the initial work with the nitrogen mustards, antimetabolites were synthesized and a number of alkaloids and antibiotics extracted from various plant, fungus, and microbial sources became available for clinical use. DeVita and colleagues introduced the first highly effective combination chemotherapy, MOPP (mechlorethamine [nitrogen mustard], vincristine [Oncovin], procarbazine, and prednisone), based on experimental studies indicating the desirability of combining agents with non-overlapping toxicities.²² Combination chemotherapy extended the curative potential for cHL to advanced disease. The ABVD (doxorubicin [Adriamycin], bleomycin, vinblastine, dacarbazine) regimen introduced by Bonadonna and colleagues represented another major advance.²³ Based on a more favorable safety profile and greater efficacy, ABVD replaced MOPP.

The estimated incidence of Hodgkin lymphoma in the United States was 9190 cases in 2014, with equal incidences in Americans of European and African descent (2.9/100,000).²⁴ The disease has a median age of onset of 38 years with bimodal incidence peaks at ages 15 to 34 and older than age 60 years (Fig. 97-1).²⁵ The second incidence peak is smaller in Americans of European descent whereas it is more prominent in Americans of Hispanic descent.²⁴ Except for Americans of Asian descent, for whom the incidence has increased by 5.2 percent per year, the incidence of Hodgkin lymphoma has been stable in the United States from 1975 to 2011 (Fig. 97-2). The nodular sclerosis subtype predominates in young adults, whereas the mixed cellularity subtype is more common in the pediatric population and at older ages. There is a male predominance at all ages (~1.4:1).

Early studies associated an increased risk of Hodgkin lymphoma in the young adult population with high socioeconomic status.²⁶ Living in a rental home, sharing a bedroom, and attending daycare or nursery school and early parity in women have been associated with reduced risk. The relationship of incidence to neighborhood socioeconomic status was demonstrated in California for younger but not for older patients.²⁵ Although associations with occupational exposure to pesticides and lifestyle factors such as cigarette smoking have been reported, the aggregate data do not indicate consistent causal relationships with exogenous chemicals or toxins. A personal or family history of an autoimmune disorder, particularly sarcoidosis, has been associated with an increased risk of Hodgkin lymphoma.²⁷ Shared etiologic factors with multiple sclerosis have been suggested but these are thought to be of minor importance.

Geographic patterns vary for the three major age groups: the incidence of Hodgkin lymphoma is greater in childhood in less-developed countries, whereas the incidence peaks in young adulthood and is associated with more favorable histologic subtypes in developed countries.²⁸ Presence of the Epstein-Barr virus (EBV) in Hodgkin and Reed-Sternberg cells is more common in less-developed countries and in pediatric and older adult cases. The worldwide incidence of the disease is much lower in the Asian population, whether residing in the Far East or in the United States, although the reported rate in Vancouver, Canada, among immigrants of Chinese descent was higher than among Chinese residing in Hong Kong.^29,30 Together these data suggest a complex interaction among possible socioeconomic, environmental, immunologic, genetic, and infectious factors in the incidence of Hodgkin lymphoma.

Demographic features have long supported a “hygiene hypothesis” postulating that one or more subtypes of cHL represent delayed exposure to an infectious agent. In 1966, MacMahon proposed that the first age peak in young adults was infectious in nature, whereas the second peak resulted from causes similar to other lymphomas.²⁶ As noted above, socioeconomic status correlates with the first, but not the second, peak.²⁵ Several reports of clustering of cHL at the time of diagnosis suggested the possibility of infectious transmission.³¹ The weaknesses of the retrospective methodology in these studies have been critically assessed, and further statistical analyses indicate that these likely occurred by chance alone.

A threefold increased risk of cHL in young adults is conferred by a prior history of serologically confirmed infectious mononucleosis. In addition, elevations in titers of EBV, the etiologic agent of infectious mononucleosis, have been reported in patients with cHL.^32,33 A large population study showed that people who developed the disease had abnormally high titers of EBV viral capsid antigen and early antigen in prediagnostic sera.³⁴ In two subsequent reports, a significantly increased risk of cHL after serologically verified infectious mononucleosis and limited to EBV-positive cases was reported in young adults.^35,36 The median incubation time was approximately 4.1 years.

EBV genomes have been detected in 30 to 50 percent of cHL tissues in developed countries, and EBV-associated cases are more common in cases with mixed cellularity histology, Hispanic ethnicity, and patients older than the age of 60 years.^37-39 Several studies report a high incidence of EBV association, 85 to 100 percent, in pediatric cHL in which geographic, ethnic, and racial factors have been implicated in the association.^37-39 The incidence of cHL is 10 to 20 times higher in patients with HIV infection than in the general population, and such cases typically have detectable EBV within Hodgkin and Reed-Sternberg cells.⁴⁰ In contrast to non-Hodgkin lymphoma, the incidence of cHL in the HIV-infected population has increased despite less-severe immunosuppression in the era of highly active antiretroviral therapy.^41,42

Genetic susceptibility and familial aggregation appear to play a role in the incidence of cHL. The increased risk of the disease among identical, but not fraternal, twins provides the strongest evidence for a genetic association.⁴³ cHL-prone families, with or without other forms of cancer, have been described in the literature and it is estimated that 4.5 percent of cases are familial.^44-46 The standard incidence ratio for age-specific familial risk from the Swedish Cancer Registry was higher for Hodgkin lymphoma (4.8) than any other neoplasm.⁴⁷ The relative risk for familial disease is stronger in individuals older than 40 years of age, males, and siblings, and a shared risk with chronic lymphocytic leukemia and non-Hodgkin lymphoma has been described.⁴⁶ An increased incidence in same sex siblings (eight- to 12-fold) versus opposite-sex siblings (1.3- to 1.4-fold) detected in the Swedish registry is consistent with older data and has been interpreted to be supportive of an environmental influence or a pseudoautosomal susceptibility gene located on a sex chromosome.^48-50

Immunoregulatory genes within or near the major histocompatibility complex that may govern susceptibility to viral infections have been postulated to influence susceptibility to cHL, an hypothesis that is supported by the demonstration of lifelong, depressed cellular immunity in cHL patients and their healthy relatives.⁵¹ Several groups described specific human leukocyte antigen (HLA) susceptibility or resistance regions, but these data have been relatively weak and sometimes inconsistent.⁵²

ORIGIN OF THE REED-STERNBERG CELL

The histologic diagnosis of cHL is based on the recognition of the Reed-Sternberg cell in an appropriate cellular background. The classic Reed-Sternberg cell has a bilobed nucleus with prominent eosinophilic nucleoli separated by a clear space from the thickened nuclear membrane (Fig. 97-3; see also Chap. 96, Fig. 96–35). Mononuclear variants (Hodgkin cells) have similar nuclear characteristics and may represent Reed-Sternberg cells cut in a plane that shows only one lobe of the nucleus. Reed-Sternberg cells are not pathognomonic for cHL; they may be seen in reactive and other neoplastic conditions. Study of the Reed-Sternberg cell has been complicated by the fact that the neoplastic cells are sparsely interspersed among a reactive, mixed, nonclonal population of lymphocytes, eosinophils, histiocytes, plasma cells, and neutrophils. Difficulty characterizing the neoplastic cells, which account for only 1 to 2 percent of the cellular composition, led to controversy regarding the etiology and pathogenesis of cHL for more than 150 years. Molecular analyses of single cells obtained by microdissection led to the discovery that cHL and NLPHL are both clonal disorders derived from germinal center B cells, in most cases.⁵³ The need to survive negative selection in the germinal center, the determination of genetic alterations and constitutive activity of key signaling pathways, and the involvement of EBV in a subset of cases have led to hypotheses concerning the malignant transformation events leading to the formation of Hodgkin/Reed-Sternberg cells. Application of additional genomic technology promises to clarify the molecular changes underlying malignant transformation and cellular proliferation.

Antigen Receptor Rearrangements

Reed-Sternberg cells and their mononuclear variants demonstrate inconsistent lineage-specific antigen expression that is unlike any other cell of the hematopoietic system. The origin of these cells was eventually determined through isolation of single cells by micromanipulation of histologic sections and analysis for immunoglobulin variable gene rearrangements.^53,54 Nearly all Hodgkin and Reed-Sternberg cells have rearranged and somatically mutated immunoglobulin VH genes, indicating a germinal center or postgerminal center origin of classic Hodgkin and Reed-Sternberg cells.^55-57 Extrapolating from the fact that a subset of these cells carries crippling mutations, it is possible that Hodgkin and Reed-Sternberg cells originate from a preapoptotic germinal center B cell with unfavorable mutations that has escaped negative selection. Rare cases of cHL with a clonal T-cell receptor gene rearrangement have been observed.⁵⁸ In contrast, single-cell analyses of NLPHL demonstrated clonal immunoglobulin gene rearrangements with ongoing mutations, an intraclonal diversity consistent with a germinal center origin of lymphocyte and histiocytic cells.^59-61

Reprogramming of Hodgkin and Reed-Sternberg Cells

Hodgkin and Reed-Sternberg cells show a global loss of their B-cell phenotype, retaining only B-cell features associated with their interaction with T cells and their antigen-presenting function.⁶² Furthermore, Hodgkin and Reed-Sternberg cells express markers of other lineages, including T cells, dendritic cells, cytotoxic cells, and myeloid cells. The lack of expression of numerous B-cell genes is the result of loss of transcription factor expression (OCT2, BOB1, PU.1) and epigenetic silencing.^63-65 The main B-cell lineage commitment factor, PAX5, is typically expressed, but its target genes are downregulated.^66,67 Reduced expression of target genes likely reflects the fact that B-cell genes are regulated by coordinated action of multiple transcription factors.

The heterogeneity of expression of myeloid, T-cell, dendritic cell, and other genes by Hodgkin and Reed-Sternberg cells is the result of many factors. Early B-cell factor 1 levels are low, de-repressing the expression of T-cell and myeloid genes and lowering transcription of B-cell–specific genes.⁶⁸ Notch 1, which plays a key role in promoting T-cell differentiation and inhibiting B-cell development, is expressed in Hodgkin and Reed-Sternberg cells.⁶⁹ Notch 1 also contributes to the expression of GATA2, a transcription factor required for proliferation and survival of hematopoietic stem cells.⁷⁰ The hematopoietic stem cell regulator polycomb G proteins are also expressed by Hodgkin and Reed-Sternberg cells and are thought to contribute to the expression of markers of different hematopoietic lineages.⁷¹ The signal transducer and activation of transcription factors (STAT) 5A and STAT5B are implicated in Hodgkin and Reed-Sternberg reprogramming as they upregulate CD30 and downregulate B-cell–receptor expression.⁷² Together, these factors cause a global loss of the B-cell phenotype and aberrant expression of genes of other cell lineages.

Genetic Alterations and Signaling Pathways

Because Hodgkin and Reed-Sternberg cells lack expression of functional B-cell surface receptors, rescue from apoptosis is probably an important mechanism of survival.^53,73 The most prevalent genetic lesions in Hodgkin and Reed-Sternberg cells involve two signaling pathways: Janus kinase (JAK)-STAT and nuclear factor-κB (NF-κB). Hodgkin and Reed-Sternberg cells have frequent gains in JAK2 and inactivation of the negative regulator of JAK-STAT signaling, suppressor of cytokine signaling 1, resulting in enhanced cytokine signaling.^74,75 Genetic alterations in NF-κB include gains and amplifications of the NF-κB transcription factor REL in about half the cases of cHL.⁷⁶ Somatic mutations of the gene encoding the inhibitor of NF-κB (IκBα) occur in approximately 20 percent of cases.^77,78 Inactivating mutations and deletions of the gene encoding A20, a negative regulator of NF-κB, have been found in approximately 40 percent of cases, nearly all of which were EBV-negative.⁷⁹

Autocrine and paracrine signaling events also contribute to constitutive activation of the JAK-STAT pathway and NF-κB transcription.⁷² STAT factors are activated by autocrine means through expression of interleukins (ILs) 13 and 21 and their receptors by Hodgkin and Reed-Sternberg cells and augmented by NF-κB activity.^80-82 Receptor tyrosine kinases expressed in these cells may also contribute to STAT activation. The tumor necrosis factor receptor family, which includes CD30, CD40, transmembrane activator, calcium modulator, and cyclophilin ligand interactor (TACI), B-cell maturation antigen (BCMA), and receptor activator of NF-κB (RANK), is involved in NF-κB signaling through interactions with the cHL microenvironment or in an autocrine fashion.^83,84

Multiple receptor tyrosine kinases are aberrantly expressed in Hodgkin and Reed-Sternberg cells, including platelet-derived growth factor receptor-α. In addition, deregulated and constitutive activation of the phosphoinositide 3′-kinase (PI3K)-AKT and extracellular signal-regulated kinase (ERK) pathways are implicated in Hodgkin and Reed-Sternberg cells. The activator protein 1 (AP1) transcription factors also appear to play a role, inducing target genes such as galectin 1 and CD30 in Hodgkin and Reed-Sternberg cells.

Several factors point to the pathogenetic role of EBV in approximately 40 percent of classical cHL. The viral proteins latent membrane protein 1 (LMP1) and latent membrane protein 2 (LMP2), in particular, appear to have hijacked signaling pathways to promote the survival of EBV-infected Hodgkin and Reed-Sternberg cells. LMP1 induces constitutive NF-κB signaling by mimicking the CD40 receptor and can activate JAK-STAT, PI3K, and AP1 signaling. LMP2 functions as a surrogate for the B-cell receptor. The role of EBV in the pathogenesis of cHL also is supported by the findings that (1) there is an inverse relationship between expression of multiple receptor tyrosine kinases and EBV expression, (2) there is an ability of EBV to rescue crippled germinal center B cells in the laboratory, (3) mutations preventing any B-cell receptor expression are in EBV-positive Hodgkin and Reed-Sternberg cells, and (4) there is an inverse relationship between mutations reducing the expression of the NF-κB regulator A20 and EBV-positive Hodgkin and Reed-Sternberg cells.

Overall, genetic alterations involving the JAK-STAT and NF-κB signaling pathways and further activation via autocrine or paracrine mechanisms interact to support the growth and survival of cHL cells. In the EBV-positive subset of patients, viral genes can provide the pathogenetic function of genetic lesions found in EBV-negative cases.

ROLE OF THE MICROENVIRONMENT

The survival of Hodgkin and Reed-Sternberg cells appear to be dependent on their microenvironment, which represents 95 to 99 percent of the cellular composition of the tumor. Hodgkin and Reed-Sternberg cells attract T cells, B cells, neutrophils, plasma cells, eosinophils and mast cells by secreting chemokines (Fig. 97-4). For instance, CCL5, CCL17, and CCL22 attract T-helper 2 and T-regulatory cells. Other chemokines attract eosinophils and mast cells, and IL-8 attracts neutrophils. These chemokines may also have direct effects on Hodgkin and Reed-Sternberg cells. T cells represent the largest and probably most important population. CD4+ T cells trigger CD40 signaling and CD4+ T-regulatory cells have potent immunosuppressive activity against infiltrating cytotoxic T cells. Other interactions include activation of TACI and BCMA through their ligand production by neutrophils and activation of CD30 through CD30 ligand-expressing mast cells and eosinophils. Connective tissue cells and their products can be involved in complex interactions, such as the stimulation of fibroblasts via factors expressed by Hodgkin and Reed-Sternberg cells and the consequent secretion by these fibroblasts of eotaxin and CCL5, which attract eosinophils and T-regulatory cells to the cHL microenvironment.

The differentiation of CD4+ T cells to T-regulatory cells and the immunosuppressive features of the cHL microenvironment have received much attention. A hallmark of these changes is the shift from an antitumor, cytotoxic T-helper 1 response to a protumor, humoral T-helper 2 response. Hodgkin and Reed-Sternberg cells produce a number of immunosuppressive factors such as IL-10, transforming growth factor-β, galectin 1, and prostaglandin E₂. Hodgkin and Reed-Sternberg cells express programmed cell death protein 1 (PD1) ligand that binds and inhibits T-cell cytotoxic function.

PRESENTING MANIFESTATIONS

History and Physical Examination

Constitutional symptoms accompany the diagnosis of cHL in approximately 30 percent of cases. Fever in excess of 38°C, drenching night sweats, and weight loss exceeding 10 percent of baseline body weight during the 6 months preceding diagnosis are designated as symptomatic “B” disease. Fevers are usually of low grade and irregular. Rarely, a cyclic pattern of high fevers for 1 to 2 weeks alternating with afebrile periods of similar duration, known as Pel-Ebstein fever, is present at diagnosis and is virtually diagnostic of the disease.^85,86 Generalized pruritus, often accompanied by marked excoriation, may be present at diagnosis; but does not confer prognostic significance. Pain in involved lymph nodes immediately after the ingestion of alcohol is a curious complaint that occurs in fewer than 10 percent of patients but is nearly specific to cHL.⁸⁷ The etiology of these symptoms has been the subject of speculation, but remains largely unexplained. Patients with extensive intrathoracic disease may present with cough, chest pain, dyspnea, and, rarely, hemoptysis. Infrequently, patients present with bone pain, including the constellation of back pain accompanied by signs and symptoms of spinal cord compression.

Detection of an unusual mass or swelling in the superficial, supradiaphragmatic lymph nodes (60 to 70 percent cervical and supraclavicular, 15 to 20 percent axillary) is the most common presentation of cHL. Only 15 to 20 percent of patients have subdiaphragmatic disease at presentation.⁸⁸ Lymphadenopathy is usually nontender and has a “rubbery” consistency. By inspection, a diffuse, puffy swelling rather than a discrete mass may be apparent in the supraclavicular, infraclavicular, or anterior chest wall regions. Infrequently, compression of the superior vena cava will result in facial swelling and engorgement of the veins in the neck and upper chest. Auscultation of the chest may reveal a pleural effusion. Rarely, a significant pericardial effusion is present at diagnosis. Palpation of the abdomen may reveal intraabdominal masses or hepatosplenomegaly, although physical examination is relatively insensitive for detection of these abnormalities.

Paraneoplastic Findings

A number of rare paraneoplastic syndromes have been described in cHL at the time of diagnosis. These include “vanishing bile duct syndrome” and idiopathic cholangitis with clinical jaundice, the nephrotic syndrome with anasarca, autoimmune hematologic disorders (e.g., immune thrombocytopenia or hemolytic anemia), and neurologic signs and symptoms.^89-91 Although parenchymal involvement of the central nervous system or meningeal involvement is rare in cHL, paraneoplastic syndromes include subacute cerebellar degeneration, myelopathy, progressive multifocal encephalopathy, and limbic encephalitis.^91,92

RADIOGRAPHIC FEATURES

Intrathoracic disease is present at diagnosis in two-thirds of patients. Mediastinal adenopathy is common in cHL, particularly in young women with the nodular sclerosis subtype.⁹³ Hilar adenopathy, pulmonary parenchymal involvement, pleural effusions, pericardial effusions, and chest wall masses may be appreciated by chest computed tomography (CT); these are more common in the presence of extensive mediastinal disease. CT of the abdomen and pelvis is routinely employed in the diagnostic evaluation of cHL. Although technologic advances have greatly increased the resolution of this technique and the subsequent detection of celiac, portal, splenic hilar, and mesenteric lymph nodes, the correlation with histologic involvement of the spleen, historically determined by laparotomy staging, has been disappointing.

Whole-body ¹⁸F-fluorodeoxyglucose positron emission tomography (FDG-PET) has become standard in the staging of cHL (Fig. 97-5).^94,95 FDG-PET correlates well with CT evaluation and may demonstrate additional areas of disease, although this information uncommonly results in changes in stage or choice of initial therapy.^95,96 FDG-PET, however, is more sensitive to bone and hepatic disease and a diffuse increase in signal can be seen at diagnosis in patients with neutrophilia. FDG-PET imaging is superior to CT scanning in distinguishing active residual disease (increased glucose metabolism) from inactive residual tissue, a major problem in assessing remission status after treatment, and has been incorporated in formal revised response guidelines.^94,95 False-positive FDG-PET scans can be seen in the marrow during or at the end of treatment as a consequence of the chemotherapy effect or use of hematopoietic colony-stimulating factors. In followup, false-positive studies may be caused by thymic hyperplasia, granulomatous disease, or infectious disorders. In addition to evaluation of residual masses, FDG-PET has been incorporated in early response monitoring for risk stratification and, in clinical trials, to alter therapy.^96-102 The predictive accuracy of FDG-PET is dependent on expertise of the imaging staff and clinical correlation. In most situations, but particularly in FDG-PET, avid anatomic sites that were previously uninvolved or those without concomitant abnormality on CT scan, usually require tissue biopsy confirmation. Combined CT and FDG-PET technology is now standard for staging and posttreatment response evaluation and has resulted in improved anatomic definition of sites with increased signal.^94-96

NODULAR LYMPHOCYTE PREDOMINANCE SUBTYPE

There is a strong correlation between age at onset, the anatomic extent of disease and histologic subtype of Hodgkin lymphoma. Approximately 5 to 10 percent of patients present with NLPHL, which is considered to be a distinct subtype, discrete from cHL.^103,104 Progressive transformation of germinal centers may precede or follow NLPHL in other sites. The cellular composition is predominantly benign B lymphocytes with or without histiocytes. The characteristic multilobated, CD20+ lymphocyte and histiocytic cells are relatively abundant (see Chap. 96, Fig. 96–40). Patients most commonly present with stage I disease (70 percent) in peripheral lymph node sites, particularly in the axillae, and there is a 3 to 4:1 male predominance.^103,104 NLPHL may be associated with diffuse large B-cell non-Hodgkin lymphoma as a composite tumor or a large cell lymphoma may develop at a later date.^105,106 The large cell variant, T-cell–rich B-cell lymphoma, may be difficult to distinguish from NLPHL and may occur concurrently or subsequently (Chap. 98).¹⁰⁷

NODULAR SCLEROSIS SUBTYPE

cHL is characterized by Hodgkin and Reed-Sternberg cells that express CD30 and CD15+ but usually not typical B-cell surface markers such as CD20 or CD79B. The nodular sclerosis subtype constitutes 40 to 70 percent of cHL and is distinguished by its distinctive clinical and histologic features. This subtype typically involves lower cervical, supraclavicular, and mediastinal lymph nodes in adolescents and young adults, particularly females. Approximately 70 percent of patients with the nodular sclerosis subtype present with limited stage disease. A characteristic histologic feature is the lacunar cell, a Reed-Sternberg variant that results from retraction of the cytoplasm of Hodgkin and Reed-Sternberg cells during formalin fixation (see Chap. 96, Fig. 96–38). Another typical feature is the thickened capsule and fibrous bands that divide the lymphoid tissue into cellular nodules.

MIXED CELLULARITY SUBTYPE

Mixed cellularity cHL involves both pediatric and older age groups and is more commonly associated with advanced stage disease, constitutional symptoms, and immunodeficiency. Approximately 30 to 50 percent of patients present with this histology. Classic Hodgkin and Reed-Sternberg cells are easily found amid a cellular background composed of lymphocytes, eosinophils, plasma cells, and histiocytes (see Chap. 96, Fig. 96–39). A worse prognosis has been reported for this subtype.¹⁰⁸

LYMPHOCYTE-DEPLETED SUBTYPE

Lymphocyte-depleted cHL has two morphologic variants: reticular and diffuse fibrosis. The reticular variant contains abundant pleomorphic neoplastic cells whereas the diffuse fibrosis variant has a prominent fibroblastic proliferation with few normal lymphocytes. Hodgkin and Reed-Sternberg cells are sparse. The lymphocyte-depleted subtype presents in older patients with symptomatic, extensive disease, and may be associated with fever of unknown origin, jaundice, hepatosplenomegaly, or pancytopenia. Peripheral and mediastinal adenopathy is much less common than in other subtypes. Cases of cHL developing in the setting of HIV typically exhibit the lymphocyte-depleted morphology.

LYMPHOCYTE-RICH SUBTYPE

The lymphocyte-rich subtype of cHL was introduced by the World Health Organization classification in 1999 (see Table 97–1) following an expert pathology review of cases of NLPHL.¹⁰⁹ The two subtypes differ subtly on morphologic grounds, but the major difference is that the Hodgkin and Reed-Sternberg cells in lymphocyte-rich cHL exhibit the classic CD30+ CD20− immunophenotype. The presenting features are very similar although patients with the lymphocyte-rich subtype tend to be older compared with NLPHL patients.¹⁰⁹ A higher rate of multiple relapses and a more favorable prognosis upon relapse is characteristic of NLPHL.

In approximately 70 percent of patients, cHL presents in the cervical nodes; in 12 percent, in the axillary nodes; and in 9 percent, in the inguinal nodes.¹¹⁰ A small minority of patients presents exclusively with subdiaphragmatic disease. In an historical series of 285 consecutive, unselected, and untreated patients evaluated at Stanford University, involvement of abdominal lymph nodes and the spleen was documented in 272 upon laparotomy, a surgical diagnostic procedure in which the intraabdominal and pelvic lymph nodes are biopsied, the spleen is removed and examined pathologically in thin slices, the liver is biopsied by needle and wedge technique, and the marrow is biopsied. The frequency of splenic involvement at laparotomy in untreated patients averaged 37 percent in 17 published series.¹¹⁰ Involvement of the spleen was strongly dependent on histologic subtype: it was involved in 60 percent of mixed cellularity and lymphocyte-depleted cases compared with 34 percent of nodular lymphocyte-predominant and nodular sclerosis cases. Hepatic and marrow disease were invariably associated with splenic involvement.

Two different theories, the “contiguity” theory of Kaplan and Rosenberg¹¹¹ and the “susceptibility” theory of Smithers,¹¹² have been proposed for the mode of spread of cHL. In support of the former, most cases of cHL appear to spread via lymphatic channels to contiguous lymphatic structures in a predictable, nonrandom pattern. Controversy has surrounded the mode of spread to the spleen, which lacks afferent lymphatics. When four or more lymph node regions are involved, spread by hematogenous distribution appears likely. Disseminated disease is more common in mixed cellularity and lymphocyte-depleted cases, consistent with the presence of reported vascular invasion.¹¹³ Whereas vascular invasion is controversial, it is more common in the spleen than in lymph nodes and connotes a poor prognosis.¹¹⁴

Optimal management of Hodgkin lymphoma relies on knowledge of the extent of disease dissemination, which is determined by performing a careful physical examination, supplemented by laboratory tests and radiographic imaging studies (Chap. 95). This process, known as “staging,” has been refined by a series of international working groups. The current 2014 Lugano staging system⁹⁴ is a refinement of the historical Ann Arbor classification¹⁵ and assigns patients to one of four stages as indicated in Table 97–2. The classification for Hodgkin lymphoma is further refined by designating the presence or absence of constitutional “B” symptoms (fever >38°C, weight loss >10 percent of body weight, or drenching night sweats). Extranodal disease, representing extracapsular extension of lymph node disease that could be incorporated in a standard radiotherapy field, is distinguished from disseminated, stage IV disease. The correlation of this staging classification system with prognosis was extensively verified when radiotherapy served as the principal treatment for all but stage IV disease. Recommended staging procedures for untreated patients have evolved with changes in therapy. Exploratory laparotomy and splenectomy, which historically advanced about one-third of clinical stages I and II patients to pathologic stages III and IV, is no longer performed. CT of the chest, abdomen, and pelvis and FDG-PET imaging provide sensitive delineation of involved sites, and the use of chemotherapy in all stages of disease has reduced the critical nature of detecting subclinical disease. Marrow involvement occurs in approximately 12 percent of new patients and is more common in patients of older age, advanced stage, less-favorable histology, or those with constitutional symptoms or immunodeficiency. Because the marrow is almost never involved in young, asymptomatic patients with favorable clinical stage I or II presentations, marrow biopsy may be omitted in staging such patients.

There are no diagnostic laboratory features of cHL. A complete blood count may reveal granulocytosis, eosinophilia, lymphocytopenia, thrombocytosis, or anemia. Anemia is usually the result of “chronic disease,” but rarely may be caused by hemolysis associated with a positive direct antiglobulin (Coombs) test. Thrombocytopenia may occur as a result of marrow involvement, hypersplenism or an immune mechanism. Immune neutropenia can occur in cHL. Cytopenias are particularly common in advanced-stage disease and the lymphocyte-depleted subtype. Elevation of the erythrocyte sedimentation rate (ESR) is most common in advanced disease and correlates with constitutional symptoms.^115,116 The degree of sedimentation rate elevation correlates with prognosis, particularly in limited-stage disease.¹¹⁷ Although nonspecific, it may be a useful harbinger of recurrent disease if serially monitored. Serum lactate dehydrogenase levels are elevated in 35 percent of patients at diagnosis.^118,119 The alkaline phosphatase may be elevated in cHL, nonspecifically in limited disease, or in association with involvement of liver, bone, or marrow in advanced disease. Hypercalcemia is unusual in cHL and when present may be secondary to synthesis of increased levels of 1,25-dihydroxyvitamin D by Hodgkin lymphoma cells.¹²⁰ A variety of other abnormalities have been reported, including hypoglycemia^121,122 resulting from autoantibodies to insulin receptors and hyponatremia resulting from inappropriate secretion of antidiuretic hormone.¹²³

Anemia, granulocytosis, lymphopenia, and low serum albumin constitute four of seven adverse prognostic factors identified in advanced Hodgkin lymphoma by an international consortium.¹²⁴ Similar to the non-Hodgkin lymphomas, serum β₂-microglobulin levels correlate with tumor burden and prognosis in Hodgkin lymphoma.¹²⁵ Serum levels of cytokines including soluble CD30, IL-6, IL-10, and the IL-2 receptor, have been reported to correlate with constitutional symptoms and advanced disease.^126-129 Examination of pleural fluid in Hodgkin lymphoma may reveal transudative, exudative or chylous characteristics. Because cytology rarely yields diagnostic Hodgkin and Reed-Sternberg cells, the etiology is most often considered to be one of central lymphatic obstruction. Laboratory abnormalities, including abnormal liver function tests associated with marked enlargement of porta hepatis nodes and biliary obstruction or intrahepatic cholestasis, may be prominent in rare presentations of Hodgkin lymphoma.¹³⁰ Similarly, the nephrotic syndrome, characterized by edema, azotemia, hypoalbuminemia, and hyperlipidemia, may rarely be present at the time of diagnosis of Hodgkin lymphoma.¹³¹

Clinically enlarged lymph nodes may be associated with a variety of infectious, inflammatory, autoimmune, and neoplastic disorders. Biopsy of unexplained, persistent, or recurrent adenopathy should be reviewed by an experienced hematopathologist. Distinction of cHL¹³² from primary mediastinal B-cell lymphoma may be difficult based on both clinical and histologic features. Gene expression profiling studies suggest that these disorders are closely related pathogenetically.^132-135 Mixed cellularity Hodgkin lymphoma may demonstrate varied cellular and stromal compositions and must be distinguished from peripheral T-cell lymphoma and T-cell–rich, B-cell lymphoma, which can also be difficult to distinguish from NLPHL.¹³⁶ Immune markers of Hodgkin and Reed-Sternberg cells, such as CD30, CD20, and CD15, are invaluable for differential diagnosis (see Chap. 96, Figs. 96–36 and 96–41). Nonneoplastic conditions that simulate Hodgkin lymphoma include viral infections, particularly infectious mononucleosis. Depleted nodes of any histology, including the depleted phase of lymph nodes from HIV-infected patients, may resemble the diffuse fibrosis variant of lymphocyte-depleted Hodgkin lymphoma. Histologic assessment of extranodal sites depends upon the organ involved and whether there is a known diagnosis of cHL. Identification of typical Hodgkin and Reed-Sternberg cells in needle biopsies of liver and marrow biopsies are not required because the specimens for evaluation are typically very small.

HISTORICAL PERSPECTIVE

Hodgkin lymphoma first became a curable neoplasm through the systematic study of the spread of the disease and the use of higher dose, extended field radiotherapy delivered with supervoltage techniques.¹³⁷ When used alone, radiotherapy doses to involved fields usually ranged from 3500 to 4400 cGy with prophylactic doses of 3000 to 3500 cGy to uninvolved tissues. The mantle, paraaortic region, and pelvis constitute the classic radiotherapy regions. With increased recognition of late effects, radiation therapy has been modified to reduce field size to areas of known or bulky disease and doses have been lowered, coincident with addition of systemic chemotherapy. Furthermore, initial disease reduction with chemotherapy results in less radiation exposure to the neck, female breast, heart, and lungs, all of which are anticipated to result in fewer late complications. Advances in radiotherapy techniques deliver more precise dose distributions, sparing normal tissues. The first modern combination chemotherapy program was the MOPP regimen devised by Devita and colleagues.²² The national mortality figures for cHL decreased by more than 60 percent in the decade that followed the introduction of MOPP.¹³⁸ Bonadonna and colleagues developed an important alternative regimen for the treatment of cHL. ABVD, which was effective in the treatment of patients who had failed MOPP^139,140 and offered a more favorable toxicity profile. ABVD subsequently became the preferred primary chemotherapy regimen, alone or in combination with radiotherapy.^141,142 Of the multiple alternative chemotherapy regimens introduced for the treatment of advanced cHL, only the bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, prednisone, procarbazine (BEACOPP) combination developed by Diehl and colleagues has demonstrated superior cure rates in multiple phase III studies.^143,144 Table 97–3 describes the drugs, doses, and schedules of combination chemotherapy programs effective in the management of Hodgkin lymphoma.

FAVORABLE, LIMITED-STAGE DISEASE

Favorable, limited-stage Hodgkin lymphoma is typically defined in North America as asymptomatic stage I or II supradiaphragmatic disease with no bulky sites. A more restrictive definition is used in Europe based on the number of Ann Arbor sites, ESR, age and extranodal sites, as well as bulky disease (Table 97–4).¹⁴³ Approximately 35 percent of stages I and II patients meet this more limited definition of favorable disease. For many years, extended-field (subtotal lymphoid) radiotherapy administered after staging laparotomy, was the treatment of choice for early stage, favorable Hodgkin lymphoma. A change in that standard was compelled by the observation that the overall mortality rate from other causes, particularly second cancers, exceeded deaths resulting from Hodgkin lymphoma at 15 to 20 years.¹⁴⁵ Early studies from Stanford University demonstrated that involved-field radiotherapy plus chemotherapy produced results equivalent or superior to wide-field radiotherapy.¹⁴⁶ Subsequently, several randomized trials demonstrated the superiority of involved-field radiotherapy plus anthracycline-containing chemotherapy compared to extended-field radiotherapy in early stage favorable Hodgkin lymphoma.^147,148

The next series of clinical trials were designed to test the optimal number of cycles of chemotherapy and the volume and dose of radiotherapy when both modalities are used in limited Hodgkin lymphoma. The Milan Tumor Institute documented disease control in more than 95 percent of early stage cHL patients treated with four cycles of ABVD and radiotherapy, with no advantage seen for extended- field radiotherapy compared to involved-field radiotherapy.¹⁴⁹ Similarly, no advantage to more extensive radiation in combination with chemotherapy was observed in a German Hodgkin Study Group (GHSG) study.¹⁵⁰ A comparison of two versus four cycles of ABVD chemotherapy paired with 20 Gy or 30 Gy radiotherapy was made in a four-arm trial conducted by the GHSG, evaluating patients with “favorable” early stage cHL, defined as patients with two or fewer sites of disease, no masses larger than 10 cm, a normal ESR, and no extranodal sites of disease. This trial produced equivalent outstanding outcomes for all four arms, with 91 to 93 percent freedom from treatment failure after 5 years, and established two cycles of ABVD followed by 20 Gy of involved field radiotherapy as a new standard of care for early stage, favorable cHL. In a subsequent study, the GHSG evaluated eliminating drugs from the ABVD regimen, randomizing patients between four cycles of ABVD, AV, ABV, or AVD plus 30 Gy involved-field radiotherapy. The AV and ABV arms were closed early because of clearly inferior outcomes, and subsequent analysis suggests that the AVD arm is also worse than the full four-drug combination.¹⁵¹

The high cure rate with current limited chemotherapy and low-dose radiotherapy creates a high standard for comparison with alternative treatment approaches. Nevertheless, considerable interest exists in devising management strategies omitting radiotherapy altogether, largely motivated by desires to avoid secondary malignancies and late cardiopulmonary complications.¹⁴⁵ This approach is particularly favored for women younger than age 30 years, who have a very high risk of developing breast cancer if treated with mediastinal radiotherapy.^145,152 Canellos and colleagues treated 71 patients with early stage, favorable cHL with six cycles of ABVD without radiotherapy and achieved a 5-year failure-free survival of 92 percent.¹⁵³ A single institution study of ABVD versus ABVD plus radiotherapy demonstrated no significant progression-free survival difference between the treatment arms, but this trial accrued relatively small numbers of patients.¹⁵⁴ A phase III North American study assigned 405 patients with previously untreated stage IA or IIA nonbulky Hodgkin lymphoma to treatment with ABVD alone for four to six cycles or subtotal nodal radiation therapy, with or without ABVD. Among those assigned to subtotal nodal radiation therapy, patients who had a favorable risk profile received subtotal nodal radiation therapy alone and patients with an unfavorable risk profile received two cycles of ABVD plus subtotal nodal radiation therapy. After a median followup of 11.3 years, the overall survival (OS) was 94 percent among those receiving ABVD alone, compared to 87 percent for those receiving subtotal nodal radiation therapy (p = 0.04). The rates of freedom from disease progression were 87 percent and 92 percent in the two groups, respectively (p = 0.05). The investigators concluded that treatment with ABVD therapy alone was superior because of a lower rate of death from secondary malignancies and other causes,¹⁵⁵ although critics have emphasized that the large fields and high doses of radiotherapy used in this trial are obsolete and that modern radiotherapy techniques would be anticipated to have much more favorable outcomes. In a European trial, the epirubicin, bleomycin, vinblastine, prednisone (EBVP) regimen was tested against the same chemotherapy plus 20- or 30-Gy involved-field radiotherapy.¹⁵⁶ Inferiority of the EBVP combination without radiotherapy resulted in the trial's early closure. A Cochrane meta-analysis of randomized controlled trials comparing chemotherapy alone with combined modality therapy concluded that adding radiotherapy to chemotherapy improves tumor control and OS in patients with early stage Hodgkin lymphoma.¹⁵⁷

Current studies are focused on assessing the potential role of interim FDG-PET scanning as a means of identifying patients (PET-negative) for whom radiotherapy can be omitted. The European Organization for Research and Treatment of Cancer (EORTC), the Lymphoma Study Association (LySA), and Fondazione Italiana Linfomi (FIL) H10 trial randomized 1137 patients with untreated supradiaphragmatic clinical stage I/II cHL to either standard therapy or an experimental, PET-response-adapted approach. Patients on the trial were stratified into favorable and unfavorable cohorts based on the presence or absence of adverse risk factors (age ≥50 years, more than four involved nodal areas, presence of mediastinal bulk [mediastinum-to-thorax ratio ≥0.35], or ESR ≥50 mm without B symptoms or ESR ≥30 mm with B symptoms).¹⁰¹ The favorable group was randomized to either standard therapy with three cycles of ABVD chemotherapy followed by involved nodal radiotherapy (30 Gy + 6 Gy boost), or “experimental therapy” in which patients whose interim PET scans after two cycles of ABVD were negative received two more cycles of ABVD (total of four cycles) but no radiotherapy. Patients in the experimental arm who had FDG-avid foci of disease after two cycles of ABVD were switched to the BEACOPP-escalated regimen for 2 cycles followed by involved nodal radiotherapy. An independent data monitoring committee stopped the study after a median followup of 1.1 years because of inferior results in the experimental group, with a 1-year progression-free survival of only 94 percent compared to 100 percent in patients treated on the standard arm (p = 0.017). The independent data monitoring committee mandated that all patients on the experimental arm be crossed over to receive involved nodal radiotherapy.¹⁰¹ Interestingly, the opposite conclusion was reached by investigators in England conducting a similarly designed trial, dubbed the “RAPID” trial.¹⁰² In this study, patients with early stage cHL were given three cycles of ABVD followed by a PET scan. Patients who were PET-negative after three cycles of ABVD were randomized to either receive 30 Gy of involved field radiotherapy or no further treatment. Patients who were PET-positive after three cycles received a fourth cycle of ABVD followed by involved field radiotherapy. An intent-to-treat analysis of the randomized PET-negative population revealed a 3-year progression-free survival of 94.5 percent in the combined modality group compared to 90.8 percent in those receiving only three cycles of ABVD (p = 0.23). An analysis of the data according to the treatment actually received revealed a 3-year progression-free survival of 97 percent in the combined modality group compared to 90.7 percent in the ABVD-alone group (p = 0.03). There were no statistically significant differences in OS between the groups (97.1 vs. 99.5 percent, respectively; p = 0.07). The investigators conducting this trial concluded that the inferior progression-free survival in the group receiving ABVD alone was acceptable in order to avoid late complications associated with radiotherapy. In view of the conflicting conclusions derived from these two large phase III trials, the role of interim FDG-PET/CT imaging in early stage cHL remains controversial. Nevertheless, the National Comprehensive Cancer Network has incorporated interim PET-imaging into their guidelines, suggesting that early stage cHL patients who are PET-negative after two cycles of ABVD may be treated with chemotherapy alone, without radiotherapy.¹⁵⁸

The advisability of administering bleomycin to patients over the age of 60 with classical Hodgkin Lymphoma (cHL) has been questioned due to the high risk of severe bleomycin-associated pulmonary toxicity in this age group. The German Hodgkin Disease Study Group has explored the feasibility, toxicity, and efficacy of combination chemotherapy with either ABVD or AVD in 287 early-stage favorable HL patients over the age of 60 in German Hodgkin Study Group studies HD10 and HD13.¹⁵⁹ Patients in these studies were randomized to therapy with either 2 cycles of ABVD (n = 137) or 2 cycles of AVD (n = 82), followed by involved-field radiotherapy (IFRT), or to 4 cycles of ABVD+IFRT (n = 68). Grade III–IV adverse event rates were similar in patients receiving 2 cycles of AVD and 2 cycles of ABVD (40% and 39%, respectively), but considerably higher in patients receiving 4 cycles of ABVD (65%). Bleomycin lung toxicity was rare in patients receiving 2 cycles of either ABVD or AVD but occurred in 7 of 69 patients (10%) randomized to 4 cycles of ABVD, with 3 fatal toxicities. From these findings, it was concluded that 2 cycles of bleomycin can be given safely to elderly patients with HL but that bleomycin should be omitted from subsequent cycles, in patients planned to receive more than 2 cycles of chemotherapy, due to the high risk of severe toxicity of bleomycin in this situation.

UNFAVORABLE LIMITED-STAGE HODGKIN LYMPHOMA

Patients with “unfavorable” prognostic factors (large tumor bulk defined as a mass 10 cm or larger in diameter or more than one-third of the transthoracic diameter, an ESR of 50 or greater, three or more sites of tumor involvement, the presence of B symptoms, or the presence of extranodal sites [see Table 97–4]) require more intensive treatment than do patients not exhibiting any of these features.¹⁵⁸ The EORTC and the Groupe d'Etude des Lymphomes de l'Adulte (GELA) reported results from a randomized study (H9U) in which such “unfavorable” early stage patients were randomized to treatment with either four or six cycles of ABVD or six cycles of BEACOPP, each followed by 30-Gy involved-field radiotherapy.¹⁵⁶ No significant differences were observed among these three treatment arms, establishing four cycles of ABVD and 30 Gy of involved field radiotherapy as a standard of care for these patients. The GHSG subsequently randomized 1395 patients with unfavorable early stage Hodgkin lymphoma to four cycles of either ABVD or BEACOPP_baseline followed by either 20 Gy or 30 Gy of involved field radiotherapy. In this study, four cycles of ABVD followed by 30 Gy of radiotherapy was the best approach, affording an 85 percent rate of “freedom from treatment failure” and an OS rate of approximately 95 percent after 5 years, while maintaining a favorable toxicity profile.¹⁶⁰ The administration of only 20 Gy of radiotherapy following ABVD was found to be clearly inferior to the other three treatment arms which all administered 30 Gy to this unfavorable group of patients. The subsequent GHSG HD14 study reported an advantage in progression-free survival for two cycles of escalated BEACOPP and two cycles of ABVD plus radiation therapy compared to four cycles of ABVD–radiation therapy, in a specified interim analysis, though no OS benefit was seen.¹⁶¹ After 3 years, 90 percent of ABVD–radiation therapy patients were disease-free compared to 96 percent treated with the BEACOPP–ABVD–radiation therapy regimen. The EORTC//LySA/FIL H10 trial randomized both favorable and unfavorable early stage patients to either standard therapy or experimental, PET-response adapted therapy.¹⁰¹ Of the 519 “unfavorable” patients evaluated in the interim analysis, 251 were randomized to standard ABVD for four cycles followed by 30 Gy of involved nodal radiotherapy, and 268 were randomized to PET-response adapted therapy. Patients on the experimental arm received six cycles of ABVD without any radiotherapy if the PET scan was negative after the second cycle of ABVD, whereas PET-positive patients were switched to BEACOPP_escalated for 2 cycles followed by radiotherapy. An independent data safety monitoring board stopped this trial after a median followup of only 1.1 years because of worse outcomes on the experimental, PET-response adapted arm, with 16 relapses in 268 cases compared to only nine relapses of 251 cases on the standard arm (hazard ratio of 2.4; 95 percent confidence intervals 1.4 to 4.4).¹⁰¹

ADVANCED DISEASE

ABVD became the standard therapy for advanced Hodgkin lymphoma by proving to be superior to the MOPP chemotherapy and equal to, but less toxic than, hybrid or alternating combinations with MOPP.^{141,142,162,163-165} Specifically, the incidence of secondary myelodysplasia, leukemia, and sterility was less with ABVD. The GHSG developed the BEACOPP regimen (see Table 97–3) based upon mathematical modeling that indicated a moderate increase in chemotherapy dose intensity would result in a significant increase in the cure rate. In the original HD9 study, BEACOPP was given in “standard” and “escalated” versions, the latter facilitated by granulocyte colony-stimulating factor use, and was compared with the cyclophosphamide, vincristine (Oncovin), procarbazine, prednisone (COPP)–ABVD regimen.¹⁴³ Patients with initial tumors equal to or greater than 5 cm or residual radiographic disease received 36-Gy radiotherapy after chemotherapy. The 5- and 10-year results demonstrated a significant progression-free and OS advantage for escalated BEACOPP compared to COPP–ABVD.^143,144 The cure rates in excess of 80 percent for escalated BEACOPP are the best ever recorded for a large phase III trial in advanced Hodgkin lymphoma. The superiority of escalated BEACOPP was observed regardless of the clinical international prognostic score.¹²⁴ Despite these outcomes, BEACOPP has not been universally accepted as the new standard in advanced Hodgkin lymphoma because of concerns about the acute toxicity, which includes greater need for hospitalization and transfusion, and late toxicity, which includes male and female sterility and an increased risk of secondary leukemia.^166-169 In addition, approximately two-thirds of patients received radiotherapy in the HD9 study. Two randomized clinical trials from Italy subsequently confirmed the superiority of BEACOPP over ABVD for the end point of progression-free survival.^170,171 In the first trial, BEACOPP (four escalated dose cycles plus two standard cycles) was compared to ABVD; all patients were to receive consolidation radiotherapy for bulky or residual disease. BEACOPP yielded significantly superior progression-free survival compared to ABVD although no difference in OS was observed.¹⁷⁰ In the second study, patients were randomized to ABVD versus four cycles of standard and four cycles of escalated BEACOPP with preplanned retreatment and high-dose therapy and autologous transplantation for treatment failures.¹⁷¹ This study also showed a significantly higher progression-free survival for the BEACOPP arm but no difference in OS (Table 97–5). The EORTC 20012 Trial randomized patients with “high-risk” advanced stage Hodgkin lymphoma with an international prognostic scale score of 3 to 7¹²⁴ to treatment with either BEACOPP (four escalated cycles + four standard cycles) or ABVD.¹⁷² Progression-free survival was improved in the BEACOPP arm but there was no statistically significant improvement in event-free or OS compared to ABVD.¹⁷² In each of the four cited trials, the hazard ratio for relapse with BEACOPP was approximately 0.5. Standards of care in advanced Hodgkin lymphoma continue to be debated in view of the lack of a survival benefit with the BEACOPP program, which benefits approximately 15 percent of patients while exposing 100 percent of patients to more toxicity.

Efforts to reduce the risk of treatment-related morbidity of escalated BEACOPP have included studying the combination of four cycles of escalated and four cycles of standard BEACOPP and eliminating radiation therapy. In the GHSG HD12 trial, four standard cycles of BEACOPP plus four escalated cycles performed similarly to eight cycles of escalated BEACOPP, with freedom from treatment failure rates of 85 percent and 86 percent, respectively, 5 years after treatment. In this study, there was a slight trend for worse failure-free survival in patients with residual radiographic masses on end-of-treatment CT imaging who did not received consolidative radiotherapy (90 percent after 5 years with radiotherapy and 87 percent without it, p = 0.08).¹⁷⁵ The subsequent HD15 trial randomized 2182 patients with newly diagnosed advanced stage cHL to receive either eight cycles of BEACOPP_escalated, six cycles of BEACOPP_escalated, or eight cycles of the BEACOPP-14 regimen. Patients with a persistent mass after chemotherapy measuring 2.5 cm or larger that was FDG-avid on an end-of-treatment FDG-PET imaging received consolidative radiotherapy (30 Gy). Freedom from treatment failure was similar for all three treatment arms but 5-year OS was significantly better for patients receiving six cycles of BEACOPP_escalated than for 8 cycles of BEACOPP_escalated (91.9 percent vs. 95.3 percent) as a result of higher treatment-related mortality and secondary malignancies with the latter regimen.¹⁷⁶

Numerous investigations are underway to determine if interim PET scans can direct more-intensive treatment with escalated BEACOPP to the subgroup that benefits.^{95,97,100,177,178} Likewise, interim and end-of-treatment PET scans are being used to direct the use of consolidative radiotherapy.^95,177-180 Although there is great enthusiasm regarding this approach to direct subsequent treatment, it is notable that the majority of patients remaining PET-positive after four cycles of escalated BEACOPP appear to be cured by the planned therapy, in contrast to findings with ABVD.¹⁸⁰

The Stanford group took an alternate approach to advanced Hodgkin lymphoma by abbreviating the duration of therapy and reducing cumulative drug doses in the Stanford V regimen.¹⁸¹ This approach appeared highly successful in institutional and phase II cooperative group trials; however, three randomized controlled trials have failed to show improved progression-free survival compared to standard ABVD chemotherapy.^182,183,174 The impressive efficacy of brentuximab vedotin in relapsed and refractory cHL^184,185 has also led to investigations into its incorporation into front-line regimens. This drug consists of an anti-CD30 monoclonal antibody conjugated to a highly potent antitubulin cytotoxic drug, monomethyl auristatin E. Phase I studies revealed unacceptable pulmonary toxicity when brentuximab vedotin was incorporated into the full ABVD regimen,¹⁸⁶ presumably as a result of augmenting the pulmonary toxicity of bleomycin. Omission of bleomycin, however, allows safe and effective combination therapy with brentuximab vedotin concurrently with AVD, with 96 percent of patients achieving complete remission.¹⁸⁶ An international phase III randomized comparison of standard ABVD versus brentuximab vedotin + AVD is underway for patients with advanced cHL.

The use of radiotherapy as a consolidation to combination chemotherapy in advanced cHL is controversial. Encouraging data from single institutions in adults and children were not borne out in randomized trials, some of which were criticized as underpowered. Furthermore, chemotherapy regimens have evolved over the time span of these studies. The application of 30 Gy involved-field radiotherapy to patients in complete remission after MOPP–ABV was studied in an adequately powered phase III trial.¹⁸⁷ No significant difference in failure-free survival was observed. Of note, all patients in partial remission received 40 Gy on this study and their outcome was not different from complete remission patients. The GHSG HD12 study randomized patients to observation or consolidation radiotherapy, with the incorporation of a central review panel, following BEACOPP. The final analysis of this study reported a slight decrease in the freedom from treatment failure if radiotherapy was omitted in patients who had persistent radiographic abnormalities on CT imaging after chemotherapy (90.4 percent vs. 87 percent after 5 years), but no difference was seen in patients with bulk disease in complete response after chemotherapy. Unfortunately, FDG-PET imaging was not routinely used in this trial.¹⁷⁵ The two Italian studies comparing ABVD with BEACOPP mentioned above routinely incorporated consolidation radiation therapy for residual or bulky disease.^170,188 The HD15 study limited the use of radiotherapy to patients with positive PET scans after four cycles of chemotherapy. With this approach, only 12 percent of patients received radiotherapy and the progression-free survival rate among PET-negative patients, who did not receive radiotherapy, was 96 percent after 1 year.¹⁷⁶ In sum, the data do not support the routine use of radiotherapy following a full course of chemotherapy in advanced Hodgkin lymphoma. However, the role of this potent treatment in patients with an early incomplete response or following a brief course of chemotherapy is likely to be more significant.^180,181

NODULAR LYMPHOCYTE PREDOMINANT HODGKIN LYMPHOMA

NLPHL presents as asymptomatic, limited-stage disease in most (~80 percent) patients.^103,104,109 Peripheral lymph nodes in the neck, axilla, or groin are commonly involved as stage IA disease. The European Task Force on lymphoma reported a 96 percent complete response rate and 99 percent and 94 percent 8-year disease-specific survival for stages I and II disease, respectively.¹⁰⁹ Because of the low likelihood of occult disease in nodular lymphocyte-predominant Hodgkin lymphoma and the tendency for the disease to remain localized for years, regional radiation therapy is considered the treatment of choice for early stage disease.¹⁰⁴ Analyses from the GHSG demonstrate that outcomes with limited radiation therapy are comparable to the use of more extensive radiation and combined modality regimens.¹⁸⁹ For the 20 percent of patients who present with stage III or IV disease, most authorities advise treatment with ABVD-based chemotherapy following the paradigms developed for cHL,¹⁰⁴ although some contend that alkylator-based regimens such as rituximab, cyclophosphamide, hydroxydaunorubicin, vincristine (Oncovin), prednisone (R-CHOP) may be equally effective and less toxic.¹⁹⁰ Because of the consistent high level expression of the CD20 antigen on the surface of NLPHL cells, the monoclonal antibody rituximab has been tested in this entity. Initial studies were conducted using single-agent rituximab in relapsed and refractory cases and demonstrated response rates of 94 to 100 percent with median durations of remission of 33 to 60 months, with longer remissions observed when maintenance rituximab was used.^191,192 Based on these findings,, and the low toxicity of rituximab, some authorities have extrapolated its use to the frontline setting,¹⁹⁰ adding it to radiotherapy for stage I or II patients and to ABVD for stage III or IV disease, although few data been published concerning the frontline use of this agent in NLPHL.

RECURRENT DISEASE

Historically, patients with cHL who relapsed after a full course of chemotherapy had a low chance for cure with second-line treatment, with the duration of initial remission a significant predictor of subsequent response and relapse-free survival. High-dose therapy and autologous blood stem cell transplantation improved the outlook for such patients and is routinely employed in first relapse for most patients younger than age 65 years, based on institutional and phase III trial experience.^193,194 Cure rates with transplantation range from 40 to 60 percent with transplant-related mortality less than 5 percent.^195,196 High-dose regimens include BEAM (carmustine, etoposide, cytarabine, melphalan), CBV (cyclophosphamide, carmustine, etoposide), and total-body irradiation with cyclophosphamide and etoposide. Consolidation radiotherapy is often employed to sites of pretransplantation bulk disease. The superiority of any single transplant conditioning regimen has not been definitively established; however, the use of high-dose sequential therapy coupled with tandem autologous transplantation is being tested in randomized trials.^197,198 In most cases, second-line chemotherapy with ICE (ifosfamide, carboplatin, etoposide), GVD (gemcitabine, vinorelbine, liposomal doxorubicin), DHAP (dexamethasone, cytarabine, cisplatin), or brentuximab vedotin is used to achieve a minimal disease state prior to stem cell mobilization and transplantation.^199-201 Treatment failures following autologous transplantation present a challenge, with longevity directly related to the time to relapse after transplantation. Allogeneic transplantation in multiply recurrent Hodgkin lymphoma has been limited by significant transplant-related mortality, although long-term disease control has been observed in a small subset together with anecdotal evidence of a graft-versus-Hodgkin antitumor effect. Nonmyeloablative transplantation conditioning regimens reduce transplant-related mortality but disease recurrence continues to present a major challenge, with failure-free survivals in the 20 to 30 percent range.^201-203

As mentioned above, the anti-CD20 antibody rituximab achieves high response rates in relapsed NLPHL and can be used as retreatment or as an extended-treatment regimen.^191,204 Monoclonal antibodies directed against the CD30 antigen are well tolerated in classical Hodgkin lymphoma but have limited therapeutic value.²⁰⁵ However, the anti-CD30 antibody–drug conjugate, brentuximab vedotin has major activity with 96 of 102 patients with relapsed or refractory cHL experiencing tumor shrinkage in a phase II clinical trial, including 75 percent with objective remissions and 34 percent with complete remissions.¹⁸⁵ The agent has also shown major efficacy when used as a bridge to allogeneic transplantation²⁰⁶ or in patients relapsing after allogeneic stem cell transplantation.²⁰⁷ The major toxicity of this agent when used as a single agent is peripheral neuropathy, though mild to moderate cytopenias also occur. Nivolumab and pembrolizumab are PD-1 blocking antibodies that have recently been shown to have remarkable efficacy in patients with relapsed or refractory cHL, with objective remission rates of up to 87 percent observed in trials enrolling heavily pretreated patients.²⁰⁸

The goal of treatment for Hodgkin lymphoma is to cure the greatest number of patients with the fewest complications. Improvements in management have resulted in cure for a large majority of patients younger than age 65 years. Survival expectations at 10 years for patients diagnosed from 2006 to 2010 exceed 90 percent for patients to age 44 years, 80 percent for patients to age 54 years, and 70 percent for patients to age 64 years.²⁰⁹ These outstanding results have been achieved by refining the use of radiotherapy and chemotherapy in the frontline setting and the development of improved secondary treatments for patients with persistent or relapsed disease. However, the late effects of treatment for Hodgkin lymphoma remain a concern for cured patients, and a small subset of patients has refractory disease.

CLINICAL PROGNOSTIC FACTORS

A number of complex prognostic factor schemes have been developed for limited Hodgkin lymphoma treated with radiotherapy alone (see Table 97–4). Massive mediastinal disease and constitutional symptoms have been consistently identified as independent predictors of relapse, whereas only older age was predictive of inferior survival. European and Canadian investigators incorporated gender, age, ESR, number of Ann Arbor disease sites, stage, and histology into stratifications for favorable, very favorable, and unfavorable disease categories. The EORTC defines four or more nodal sites, ESR greater than 50 in asymptomatic patients or ESR greater than 40 in symptomatic patients, and histology as indicators of intermediate disease, whereas the GHSG designates any one of the following: massive mediastinal disease, extranodal disease, ESR greater than 50 if asymptomatic and greater than 30 if symptomatic, and three or more nodal sites as intermediate disease (see Table 97–4). It is important to be aware of the variable eligibility criteria when interpreting the literature in early stage Hodgkin lymphoma and to note that these clinical variables are currently used to group patients for clinical investigations. The international prognostic score, based on seven factors (see Table 97–4), is used in advanced disease.^124,210 Each factor reduced the freedom from progression by approximately 7 percent. Only 7 percent of patients were in the worst prognostic group (five to seven factors) and the freedom from progression in this subset was 42 percent at 5 years. Consensus with regard to prognostic factors promotes uniformity in clinical trial design and provides a rationale for alternate approaches in high-risk subsets. The superiority of escalated BEACOPP was seen across the spectrum of the international prognostic score, but approximately 30 percent of high-risk (four to seven factors) patients experience treatment failure.¹⁴⁴ Alternately, the improvement in progression-free survival with BEACOPP over ABVD in an Italian study was limited to higher-risk patients.¹⁷⁰ Age has been a consistent adverse prognostic marker regardless of tumor burden, but recent gains eliminate this adverse factor up to approximately age 55 years.^211,212 Although less-intensive treatment may explain inferior results in a subset of patients, results in older patients are worse, even when the intensity of therapy is controlled.²¹² The BEACOPP regimen was associated with unacceptable toxicity and demonstrated no advance over ABVD for patients older than age 65 years.²¹³

Normal results from FDG-PET imaging at the completion of treatment confers a high negative predictive value, ranging from 81 to 100 percent.¹⁷⁹ The positive predictive value at the end of chemotherapy is more variable and is related to disease extent and use of radiotherapy.^179,214,215 There is great interest in FDG-PET imaging after one to three cycles of chemotherapy, as several studies indicate that negative results predict treatment success whereas positive results predict a high likelihood of treatment failure.^95,96,98,99 Numerous ongoing phase III clinical trials are designed to alter treatment based on early PET results with a goal of achieving high cure rates and minimizing toxicity.¹⁰⁰ PET status prior to autologous transplantation has also emerged as a dominant prognostic factor.²¹⁶

Adverse prognostic factors identified for patients with relapsed cHL undergoing autologous stem cell transplantation include (1) duration of first complete remission less than 1 year, (2) failure to achieve a second complete remission with salvage chemotherapy prior to transplantation, (3) presence of B symptoms or extranodal sites of disease involvement, and (4) persistence of FDG-avidity on PET imaging prior to transplant.^216-218

Patients with primary progressive Hodgkin lymphoma have the least-favorable prognosis. Fortunately, newer treatment approaches have reduced the proportion of patients in this category.

Clinical prognostic factors are surrogates for the underlying cellular and molecular biology of Hodgkin lymphoma. Serum levels of cytokines, including soluble CD30, a probable marker of tumor burden, and IL-10, a measure of immunosuppression related to the microenvironment, are associated with adverse prognosis independent of clinical features.²¹⁹ The chemokine CCL17, secreted by Hodgkin and Reed-Sternberg cells, is elevated in patients' sera and is being studied as a marker of response.^220,221 Multiple, but not all, investigators found BCL-2 expression to be of prognostic significance.^222-225 CD20 expression by Hodgkin and Reed-Sternberg cells has been associated with less-favorable outcomes in some, but not all, series.^226,227 Lack of HLA class II expression, which may allow immune escape of Hodgkin and Reed-Sternberg cells, was found to be an independent prognostic factor.²²⁸ The prognostic significance of EBV association varies with age, conferring an adverse outcome in older individuals.^229,230 Furthermore, single nucleotide polymorphisms in HLA-A2 have been associated with risk of EBV-positive Hodgkin lymphoma.⁵² A number of studies have focused on the inflammatory microenvironment of Hodgkin lymphoma. Increased numbers of T-regulatory cells have correlated with favorable outcomes and decreased numbers of markers for cytotoxic T cells have correlated with adverse outcomes in several series.^231-233 Similarly, heavy infiltration of cHL with macrophages (detected by immunohistocytochemical staining for CD68), is associated with poor outcome, presumably because of secretion of immunosuppressive cytokines.²³⁴ Finally, a 23-gene expression signature that can be assessed in formalin-fixed paraffin embedded tissue biopsies has been shown to have major prognostic power in patients with cHL.²³⁵ Together, these findings suggest an important interplay between characteristics of Hodgkin and Reed-Sternberg cells and the inflammatory environment.

COMPLICATIONS OF TREATMENT

The treatment of Hodgkin lymphoma is associated with important acute and chronic side effects. Although the acute complications of chemotherapy and radiotherapy may be troublesome, they are relatively easily managed. Late-treatment effects in the form of sterility, second malignancy, and cardiopulmonary disease are more serious and are known to contribute to shortened longevity for cured patients.^{145,152,236,237} Excess mortality from second malignancies and cardiac disease increase with time and are currently the leading causes of death for cHL patients. As treatment has evolved, the risks of radiation-related complications has lessened but long latency periods and uncertainty regarding associations with lower doses make it difficult to predict individual risks. Recognition and understanding of these problems helps to shape primary treatment choice and facilitate optimal followup for survivors.

Acute leukemia and myelodysplasia were the initial second malignancies to be observed after successful treatment for Hodgkin lymphoma with MOPP chemotherapy.²³⁸ The risk following MOPP was proportional to the cumulative dose of alkylating agents and was associated with recurring abnormalities of chromosomes 5 and 7.^238-240 Actuarial risks of approximately 5 percent with relative risks in excess of 100 have been reported over a 7- to 10-year period with alkylating agent-based therapies. The risk of secondary leukemia is greater in patients older than age 35 years. Prognosis for secondary leukemia is poor with survivals of less than 1 year.²⁴¹ The risk of acute leukemia is significantly less after ABVD chemotherapy,¹⁴¹ although it is not absent. A large international study observed a significant reduction in excess absolute risk after 1984, presumably as a consequence of change in primary therapy.²⁴¹ However, acute leukemia may complicate Hodgkin lymphoma treatment with higher doses of etoposide and doxorubicin, such as in the BEACOPP regimen.^143,166,169 This form of leukemia tends to occur earlier and be associated with balanced translocations of chromosome 11. Patients who have received second-line therapies and autologous transplantation are at highest risk for myelodysplasia and secondary leukemia.

There is an increased relative risk of non-Hodgkin lymphomas after treatment for Hodgkin lymphoma.^242,243 These are diffuse, aggressive B-cell lymphomas that may occur early or late after treatment. There is no clear relationship to the type of primary treatment. The incidence of secondary lymphoma in a series of 5406 patients treated on GHSG protocols was 0.9 percent; prognosis was worse if lymphoma developed within 3 months of primary therapy.²⁴⁴ Although prognosis was relatively poor in this series, the data antedated the routine use of rituximab in treatment regimens. It is not clear how non-Hodgkin lymphomas relate to treatment-related immunodeficiency, predisposition to B-cell malignancy, or a shared common cell of origin. Marginal zone lymphomas with identical B-cell receptor genes also have been identified.²⁴⁵ Diffuse large B-cell lymphoma and its variants are most frequent in nodular lymphocyte-predominant Hodgkin lymphoma, where they have been found to be genetically related.²⁴⁶

An increased risk of solid cancers after treatment for Hodgkin lymphoma has long been recognized, and with time, these have emerged to account for 75 to 80 percent of all cases of second malignancy.^145,152,242 The risk is related to radiotherapy exposure, with tumors occurring in or at the edges of the radiation field. The most common solid tumors are breast, lung, and gastrointestinal malignancies. The latency for developing second cancers is an important consideration, as these typically develop after at least 10 years and continue to pose excess risk for as long as 30 years after treatment. Breast cancer is increased in women treated before age 30 years and is markedly increased in children and adolescents.^152,247-249 Case-control studies have examined the relationship of radiation dose to the breast and the risk of cancer, finding a 3.2-fold increased risk with doses greater than 4 Gy and an eightfold risk for doses greater than 40 Gy.^250,251 Elimination of routine axillary radiation, which is now standard, results in a 2.7-fold reduction in risk and it is anticipated that risks may further decline with current low-dose or nodal radiotherapy. Cofactors are important for defining the risks of second breast cancer, which are highest for women younger than age 30 years when irradiated and for those who continue to have normal menses.^252,253

Lung cancer risk is greatest among patients who are older than 45 years of age when treated. Tobacco exposure has a multiplying effect and alkylating agent exposure also contributes to risk. Among patients with chest irradiation, a tobacco history, and alkylating chemotherapy, the lung cancer risk was 49-fold higher than in patients who had none of these exposures.²⁵¹ Alkylating agent chemotherapy independently increases risk of lung cancer with dose–response associations reported in population-based studies.^250,251,254

Estimates of relative risks of cardiac mortality in Hodgkin lymphoma survivors range from 2.2- to sevenfold.^{236,237,255,256} Mediastinal radiotherapy is associated with an increased risk of cardiac disease. An increased risk of death from coronary artery disease and acute myocardial infarction has been identified in adults and children.^248,256,257 Other types of cardiac disease are often asymptomatic, including valvular disease, conduction defects, and cardiomyopathy.^258,259 The risks of radiation-related heart disease do not appear to be influenced significantly by the addition of chemotherapy. The onset of increased risk is within 5 to 10 years. As risk is associated with the dose and volume of radiotherapy and the latency is 5 to 10 or more years, the hazards associated with current lower dose and smaller fields remain to be assessed.²⁵⁵ Established cardiac risk factors, including hypertension, hypercholesterolemia, and smoking, significantly contribute to the subsequent risk of cardiac disease after therapy, offering opportunities to reduce individual risks.²⁶⁰ A British report indicated an elevated risk of cardiac mortality of 7.8-fold following ABVD alone, which rose to 12.1-fold when given with mediastinal irradiation.²⁶¹ These results provide a cautionary note, but more data are needed and these results do not address cumulative exposures that are lower with modern therapy.

Noncoronary vascular complications have been reported after neck irradiation, with associations to dose greater than 36 Gy and cofactors of hypertension, diabetes and hypercholesterolemia.^262,263 In a retrospective cohort study, the standardized incidence ratio for was 2.2 for stroke and 3.1 for transient ischemic attack.²⁶² However, it is important to note that modern approaches to Hodgkin lymphoma therapy use lower radiation doses, smaller fields, and planning techniques that limit dose inhomogeneity and hot spots commonly seen in the neck area with older techniques.

Approximately 90 percent of males are permanently sterilized by six cycles of MOPP chemotherapy.²⁶⁴ The risk is related to the cumulative dose of alkylating agents such that two to three cycles of MOPP result in azoospermia in approximately 50 percent of patients.²⁶⁵ Female fertility after alkylating agent-based treatment is related to age at treatment as well as cumulative alkylating agent dose.^266,267 The ABVD combination is associated with temporary amenorrhea and azoospermia with full recovery noted in 50 to 95 percent of patients.^268,269 A case-control study found no significant reduction in fertility among women treated with ABVD.²⁷⁰ In contrast, no men had normospermia following treatment with BEACOPP and amenorrhea occurred in more than 50 percent of women.^167,168,271 Several authors have described pregnancy outcome following treatment for Hodgkin lymphoma. No increase in birth defects or complications of pregnancy has been seen.²⁶⁷

Thyroid dysfunction is common after neck irradiation, reaching a risk of 47 percent at 26 years in the Stanford series.²⁷² Thus, patients at risk should be monitored during followup observation. Rarely, hyperthyroidism, Graves ophthalmopathy or thyroid neoplasms occur after neck radiotherapy.²⁷² Lhermitte sign, a transient complaint of an “electric shock” sensation produced by head flexion, is a common sequela of mantle radiotherapy.²⁷³ The incidence of radiation pneumonitis depends on the volume of lung irradiated and the total dose. Symptoms include cough, dyspnea, and fever. Although prospective assessment of pulmonary function demonstrates reduction of lung volumes following mantle radiotherapy, recovery is seen in 12 to 24 months and symptomatic radiation pneumonitis is unusual.^274,275

Full-dose radiation therapy interferes with normal growth and development in children. Current therapy programs use low-dose or no radiotherapy for all stages of disease. Overwhelming sepsis is a rare event in patients who have been splenectomized and treated for Hodgkin lymphoma, particularly children.^276,277 Vaccination against encapsulated organisms 10 to 14 days prior to the onset of treatment is advised. However, it must be recognized that neither vaccines nor antibiotic prophylaxis may provide adequate protection. Fatigue is commonly reported in Hodgkin lymphoma survivors and has been related to pulmonary function and peak oxygen uptake.^258,278

With the high rates of cure currently attained in the management of Hodgkin lymphoma, reduction in late effects and quality of life assume even greater importance. Patient education is essential to promote healthy behaviors to reduce modifiable risk factors. In addition, early detection and prevention strategies for second cancers and cardiac disease should be considered in high-risk patients. However, the choice and efficacy of diagnostic testing, and their optimal timing and frequency, require further study.²⁷⁹ Most of the documented late effects relate to outdated chemotherapy and radiotherapy protocols, and that modeling, as well as recent data, indicate that lesser exposures significantly reduce second cancer risk.^252,280 Modern therapies will likely further reduce the risks of late complications. It continues to be important to follow long-term survivors, and the contribution of genetic and environmental factors is an important ongoing area of inquiry.