Hodgkin Lymphoma
Nancy L. Bartlett and Kelley V. Foyil
• Estimated 9060 new cases in United States in 2012 with 1190 deaths
• U.S. age-adjusted incidence rate of 2.8 per 100,000 per year
• Higher incidence among males than females
• Highest incidence in North America and Western Europe
• Bimodal age distribution (peaks at 15 to 35 years and then again later in life)
• Hodgkin Reed-Sternberg and lymphocyte-predominant (LP) cells derive from germinal center B cells. Hodgkin Reed-Sternberg and LP cells are rare in the lymphoma tissue, and interactions with other cells in the microenvironment may play a role in the pathophysiology of the disease.
• In classic Hodgkin lymphoma (HL), Hodgkin Reed-Sternberg cells express CD30 and have lost expression of B-cell markers.
• Nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL) represents a distinct clinicopathological entity characterized by scattered large “popcorn” cells that express CD45 and CD20.
• Epstein-Barr virus may be involved in the pathogenesis of some classic HL cases, particularly in tropical areas.
• Physical examination with attention to peripheral nodes and spleen
Cervical, supraclavicular, mediastinal nodes involved in majority of patients with classic HL
• History including presence or absence of pruritus, drenching night sweats, fevers, and significant weight loss
• Laboratory evaluation to include complete blood cell count with differential, albumin, and erythrocyte sedimentation rate
• Positron emission tomography/computed tomography (PET/CT)
• Bone marrow biopsy not indicated in early-stage disease and may no longer be needed in advanced-stage disease in patients undergoing PET
• Early-stage non-bulky classic HL
ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine) chemotherapy (two to four cycles) with 20 to 30 Gy of involved-field radiation therapy (IFRT) with number of cycles and dose of radiation dictated by high-risk features
Interim PET likely to select patients who benefit most from radiation therapy
• Early-stage classic HL with bulky disease
Combined modality therapy with four to six cycles of ABVD followed by IFRT
Standard risk: six cycles of ABVD
Interim PET may select patients who benefit most from escalated BEACOPP.
Combined modality therapy with two to four cycles of chemotherapy plus IFRT
Limited data: R-CHOP (rituximab + cyclophosphamide, doxorubicin, vincristine, prednisolone), R-ABVD (rituximab + ABVD), or R-CVP (rituximab + cyclophosphamide, vincristine, prednisolone) are reasonable options.
• Salvage chemotherapy followed by autologous stem cell transplant is the standard approach, with 50% to 60% achieving long-term remissions.
• Pretransplant PET/CT is highly predictive of outcomes.
• Brentuximab vedotin, an antibody-drug conjugate, has a 75% response rate in relapsed/refractory HL.
Introduction
Nearly 90% of patients with early-stage Hodgkin lymphoma (HL) and 80% of advanced-stage patients will be cured. Survival rates have continued to increase as a result of enhanced diagnostic techniques, better staging methods, improved understanding of the importance of chemotherapy dose intensity, more effective salvage regimens, and perhaps a greater appreciation of the need to screen for late complications. Although they are indirect evidence, population-based findings of improved survivals since 1980 provide credence to the idea that limiting the number of cycles of chemotherapy and decreasing the use of radiation therapy (RT) has not resulted in worse overall outcomes for patients with HL and support increasing efforts to minimize therapy in an effort to minimize late toxicities.1 With new biological prognostic markers and novel drugs for HL in development, outcomes are likely to continue to improve.
Epidemiology and Etiology
In the United States, an estimated 9060 new cases of HL were diagnosed in 2012 with 1190 deaths.2 The U.S. age-standardized incidence rate (ASR) per 100,000 per year is 2.8, with higher incidence among males. The median age at diagnosis is 38 years, and there is a bimodal age curve in resource-rich countries, showing peaks at 15 to 35 years and then later in life. Worldwide, the incidence of HL is highest in Europe (ASR 2.0), the Americas (ASR 1.5), and the Eastern Mediterranean (ASR 1.4).3 Regions with lower ASRs include Africa (ASR 0.9), Southeast Asia (ASR 0.7), and the Western Pacific (ASR 0.4). In the United States, age-adjusted incidence rates vary among races/ethnicities, with the highest incidence in non-Hispanic whites (ASR 3.6 for males and 2.9 for females), followed by blacks (ASR 3.2 for males and 2.3 for females).2 Although the incidence rates have remained stable for whites, significant annual percentage increases between 1992 and 2009 occurred for blacks and Asian/Pacific Islanders.
HL is likely to have different causes depending on patient age at diagnosis, nationality, and Epstein-Barr virus (EBV) tumor phenotype. Epidemiological studies support both environmental and genetic factors in the etiology of HL. Multiple siblings, higher birth order, crowded living conditions, lower socioeconomic status, and exposure to daycare have all been associated with a decreased risk for HL in young adults. Historically, this has been interpreted as evidence to support late exposure to an infectious agent as an etiologic factor in the development of HL but no causal virus has been found. Additional circumstances associated with an increased risk for HL include a lower level of fecal-oral exposure early in life (e.g., lower incidence in thumb suckers), appendectomy, tonsillectomy, eczema, and smoking.4 As opposed to a specific etiologic infectious agent, others have proposed the “hygiene hypothesis” as a potential cause of HL in young adults. This theory proposes that a reduction in cumulative microbial exposures in childhood may adversely affect immune response development specifically by decreasing interleukin (IL)-12 secretion, which may lead to persistence of the T-helper (Th) type 2 phenotype, resulting in an immature immune response phenotype.4,5
EBV has been long thought to play a role in the pathogenesis of a subset of HL cases because EBV latent membrane protein (LMP-1) and/or EBV-encoded small RNAs are found in Hodgkin Reed-Sternberg cells in approximately one third of HL cases. Serum profiles of antibodies to EBV are significantly altered, both before and after the diagnosis of EBV+ classic HL.6,7 These distinctive serologic responses to EBV latent antigens are also seen in other disorders associated with immune dysfunction such as the acquired immunodeficiency syndrome, rheumatoid arthritis, and inherited immunologic disorders.6 In young adults, a history of infectious mononucleosis is a risk factor for EBV+ classic HL in some series, occurring at a median of 2.9 years after infectious mononucleosis infection.8
Support is building for a familial or inherited component in the development of HL. Mack’s landmark study of twins revealed that monozygotic but not dizygotic twins of patients with HL had a significantly increased risk for developing HL.9 Several reports from the Swedish Family-Cancer Database demonstrated that relatives of HL patients had a fourfold elevated risk for developing HL.10 If the index case was the parent, the standardized incidence ratio (SIR) was 3.05 for both sexes with the highest incidence for a son of a father with HL (SIR 3.64).11 If the affected parent was diagnosed before age 40, the SIR was 6.46. The highest risks were for male-male (SIR 8.00) and female-female (SIR 11.75) sibling pairs.
This increased risk for HL in relatives inspired genomic studies of classic HL, specifically investigating the major histocompatibility complex region (human leukocyte antigen [HLA] system). The HLA class I region is associated with EBV+ HL in numerous studies, with many focusing on the HLA-A gene. Niens and associates reported that the HLA-A*01 haplotype was associated with an increased risk for developing EBV+ HL whereas the HLA-A*02 haplotype had a decreased risk.12 These associations were confirmed in the presence of infectious mononucleosis; the authors suggested that the association between history of infectious mononucleosis and EBV+ lymphoma was not seen in the presence of HLA-A*02 because this allele appeared to “neutralize the effect of infectious mononucleosis.”13 In a genome-wide association study of classic HL and EBV, the associations of the class I region variants with EBV+ classic HL and class II region variants with EBV− classic HL were corroborated.14 Additionally, two loci in the major histocompatibility complex region were associated with classic HL, irrespective of EBV status. This suggests that there may be some components of pathogenesis that are common to both EBV− and EBV+ classic HL. Huang and associates studied HLA and EBV by using a polymerase chain reaction (PCR)-based sequence-specific oligonucleotide probe (SSOP) hybridization approach and discovered significant differences between EBV+ and EBV− classic HL cases for several probes that discriminate between HLA-A*01 and HLA-A*02.15 The authors concluded that these HLA antigens are responsible for the association of EBV with HL, not specific single nucleotide variants shared by multiple alleles.
Pathology and Biology
In classic HL, rare, scattered mononuclear (Hodgkin) and multinucleated (Reed-Sternberg) cells are found in a mixture of inflammatory and accessory cells such as eosinophils, neutrophils, histiocytes, plasma cells, and fibroblasts (Fig. 105-1A and B). T lymphocytes often ring the Hodgkin Reed-Sternberg cells in a rosette-like manner. Hodgkin Reed-Sternberg cells are uniformly positive for CD30 and frequently positive for CD15, both in a membrane pattern with Golgi staining (see Fig. 105-1C). CD20 expression is variable, MUM1 staining is positive, and PAX5 staining is weaker in Hodgkin Reed-Sternberg cells than in reactive B cells. The cells are usually negative for CD45 and CD75. The detection of EBV-encoded RNA is indicative of classic HL and is found in approximately one third of cases.
NLPHL comprises numerous large, tightly packed nodules (see Fig. 105-1D). Scattered large “popcorn” or lymphocyte-predominant (LP) cells are CD20+ but rarely CD15+ or CD30+. These LP cells are present with normal lymphocytes and histiocytes within large spherical meshworks of the follicular dendritic network. If the process is entirely diffuse, the diagnosis of T-cell–rich large B-cell lymphoma must be considered.
Both LP and Hodgkin Reed-Sternberg cells are derived from germinal center (GC) B cells.16 LP cells express several typical GC B-cell markers and grow in a follicular dendritic network. In contrast, Hodgkin Reed-Sternberg cells have lost their B-cell signature through transcriptional reprogramming, likely by DNA methylation, upregulation of NOTCH1 and other negative regulators of B cells, and the persistent activation of nuclear factor κB (NF-κB).16 Hodgkin Reed-Sternberg cells, which cannot produce immunoglobulins, should be targeted for apoptosis, but NF-κB is involved in the protection of these cells. This NF-κB protection can result from EBV infection, and these infected cells are then independent of the normal B-cell receptor survival signals.17 Other pathways with deregulated activation in Hodgkin Reed-Sternberg cells include Janus-activated kinase/signal transducer and activator of transcription (JAK/STAT), phosphatidylinositol-3-kinase/Akt (protein kinase B) (PI3K/Akt), AP-1 (activator protein-1), and mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/Erk).16,18 Tiacci and colleagues reported the first genome-wide transcriptional analysis of microdissected Hodgkin Reed-Sternberg cells in comparison with classic HL cell lines and normal B-cell subsets.19 This analysis identified two molecular subgroups of classic HL with varying strengths of NOTCH1, MYC, and IRF4 proto-oncogenes. Additionally, BCL-interacting killer (BIK), an apoptosis inducer, and inositol polyphosphate-5-phosphatase, 145kDa (INPP5D), an inhibitor of the PI3K pathway, were silenced in Hodgkin Reed-Sternberg cells.
The tumor microenvironment is critical in the development and proliferation of HL. Reactive cells assist in the proliferation of Hodgkin Reed-Sternberg cells by secreting chemokines and cytokines and creating an immunosuppressive environment. Chetaille and colleagues used gene expression profiles generated by using DNA microarrays to show that variations in the tumor environment correlate with outcome.20 Altered cytokines, chemokines, receptors, and ligands can contribute to classic HL pathogenesis by promoting the Th2 response. Examples include chemokine ligand 17 (CCL17); chemokine ligand 22 (CCL22); chemokine ligand 20 (CCL20); chemokine ligand 9 (CXCL9); chemokine receptor 3 (CXCR3); IL-13, GATA binding protein 3 (GATA3); and chemokine receptor 4 (CCR-4). The aforementioned substances can also suppress the Th1 response (ex: IL-10, transforming growth factor–β [TGF-β], programmed cell death ligand-1 [PD-1L], galectin-1) and can promote the influx of eosinophils (ex: IL-5, IL-9, eotaxin, chemokine ligand 28 [CCL28]), mast cells (IL-9, chemokine ligand 5 [CCL5]), plasma cells (IL-6), neutrophils (IL-8, granulocyte-macrophage colony-stimulating factor), and fibroblasts (IL-13, TGF-β, tumor necrosis factor–α, matrix metalloproteinase [MMP], tissue inhibitor of metalloproteinases 1 and 2 [TIMP1, TIMP2]). 21 Interactions with CD4+ T cells, a mixture of Th cells and regulatory T cells that surround the Hodgkin Reed-Sternberg cell, appear to play a critical role in Hodgkin Reed-Sternberg cell survival.18
Clinical Manifestations, Evaluation, and Staging
Common presenting symptoms include painless peripheral adenopathy (most commonly in the cervical or supraclavicular region), cough or chest pain related to mediastinal disease, and pruritus. B symptoms including recurrent drenching night sweats, recurrent unexplained fever greater than 38° C (100.4° F), and unexplained weight loss (>10% of baseline) occur in about 40% of patients. Bulky mediastinal disease is most common in young women with the nodular sclerosis subtype of classic HL (Fig. 105-2). The mixed cellularity subtype of classic HL more often is seen in older men as peripheral adenopathy or subdiaphragmatic involvement and rarely as mediastinal involvement. Lymphocyte-rich classic HL usually has a very favorable presentation with nonbulky peripheral adenopathy. Lymphocyte-depleted classic HL usually is seen at advanced stage (III to IV) as retroperitoneal lymph node, abdominal organ, and bone marrow involvement. NLPHL patients usually have limited-stage disease with cervical, axillary, or inguinal lymph node involvement.
Staging of HL is based on the Ann Arbor staging system with the “Cotswolds” modifications. Four stages document the extent of lymph node and disseminated disease (Table 105-1) with information on the presence (B) or absence (A) of B symptoms. Additionally, contiguous extranodal involvement is designated E and bulky disease, defined as either a single mass larger than 10 cm or a mediastinal mass exceeding one third of the maximum transverse transthoracic diameter measured to the inside of the ribs on a standard posteroanterior chest radiograph, is noted. Stage I HL involves a single nodal region, stage II HL involves two or more lymph node regions on the same side of the diaphragm, stage III HL involves lymph node regions on both sides of the diaphragm, and stage IV HL diffusely involves one or more extralymphatic organs or sites.
Table 105-1
Modified Ann Arbor Staging System for Hodgkin Lymphoma
| Stage | Involvement |
| I | Single lymph node region (I) or one extralymphatic site (IE).* |
| II | Two or more lymph node regions, same side of the diaphragm (II) or local extralymphatic extension plus one or more lymph node regions or same side of the diaphragm (IIE)* |
| III | Lymph node regions on both sides of the diaphragm (III), which may be accompanied by local extralymphatic extension (IIIE).* |
| IV | Diffuse involvement of one or more extralymphatic organs or sites. |
| A | No B symptoms |
| B | Presence of at least one of the following symptoms: |
| 1. Unexplained weight loss >10% baseline during 6 months before staging | |
| 2. Recurrent unexplained fever >38° C | |
| 3. Recurrent night sweats | |
| Bulky tumor† |
*E lesion: Localized extranodal extension of Hodgkin’s lymphoma from a contiguous or nearby nodal site is noted with the designation E, for example, stage IIEA for asymptomatic disease in the mediastinum with contiguous extension into nearby lung.
†Bulky tumor is defined as either a single mass of tumor tissue exceeding 10 cm in largest diameter or a mediastinal mass exceeding one third of the maximum transverse transthoracic diameter measured to the inside of the ribs on a standard posteroanterior chest radiograph.
Data from Carbone PP, Kaplan HS, Musshoff K, et al. Report of the Committee on Hodgkin’s Disease Staging Classification. Cancer Res 1971;31:1860–1.
Patients should have an excisional biopsy if possible. Core needle biopsies provide limited tissue in which to examine rare Hodgkin Reed-Sternberg cells and architecture, whereas flow cytometry rarely contributes to HL diagnosis. Pretreatment evaluation should include physical examination and history, laboratory studies (complete blood cell count, chemistries, erythrocyte sedimentation rate [ESR]), and whole-body combined positron emission tomography and computer tomography (PET/CT). Limited-stage HL is more common, with approximately 25% of patients with stage I disease, 40% with stage II disease, 20% with stage III disease, and 15% with stage IV disease.2 In addition to determining stage of disease at diagnosis, advanced-stage patients should be classified by the Hasenclever International Prognostic Score (IPS) risk group and limited-stage patients with an unfavorable or favorable status according to the German Hodgkin Study Group or European Organization for Research and Treatment of Cancer criteria (Table 105-2). Risk factor classification provides prognostic information and helps with treatment stratification.
Table 105-2
Prognostic Factors in Hodgkin Lymphoma
| Risk Factors | Early Stage | Advanced Stage | |
| European Organization for Research and Treatment of Cancer | German Hodgkin Study Group | Hasenclever | |
| Bulk | MMR > 1/3 | MMR > 1/3 | — |
| Sites | > Three nodal sites | > Two nodal sites | Stage IV |
| Extranodal sites | |||
| Laboratory | ESR ≥ 50 (if asymptomatic) | ESR ≥ 50 (if asymptomatic) | Hemoglobin < 10.5 g/dL |
| White blood cells ≥ 15,000/µL | |||
| ESR ≥ 30 (if symptomatic) | ESR ≥ 30 (if symptomatic) | Lymphocytes < 600/µL or < 8% of white blood cells | |
| Albumin < 4 g/dL | |||
| Demographics | Age ≥ 50 yr | — | Age ≥ 45 yr |
| Male gender | |||
ESR, Erythrocyte sedimentation rate; MMR, mediastinal mass ratio.
Compared with CT alone, PET/CT upstages approximately 15% of patients and downstages less than 5% of patients.22 In two series, staging PET/CT resulted in a change in treatment in 7% to 18% of patients.22,23 Traditionally, bone marrow biopsy has been recommended for patients with advanced disease. El-Galaly and colleagues reported on 454 newly diagnosed patients with HL of whom 18% had focal skeletal lesions on PET/CT and 6% had a positive bone marrow biopsy.24 No patients with stage I or II HL had a positive bone marrow biopsy, and no patients were allocated to a higher clinical risk group based on the results of bone marrow biopsy. Of 27 patients with bone marrow involvement, 23 (85%) had focal bone lesions on PET/CT. Diffuse marrow uptake was never associated with a positive bone marrow biopsy. PET/CT had a negative predictive value of 99% for bone marrow involvement. Bone marrow biopsy is unlikely to change the risk or treatment strategy in the age of PET/CT staging and can likely be eliminated in all patients.
Primary Therapy
Despite the overall excellent prognosis of the majority of patients with HL, the optimal initial treatment for both early- and advanced-stage disease remains controversial. Because of the unique efficacy of salvage therapy for HL, a slightly less effective but less toxic approach as initial therapy may ultimately optimize survivorship.25 Tailoring therapy based on results of an interim PET/CT scan performed after one to three cycles of chemotherapy may represent the optimal approach for most patients. However, mature data from randomized trials testing this concept are not yet available.
Early-Stage Nonbulky Hodgkin Lymphoma
The transition from extended-field radiation therapy (EFRT) to combined-modality therapy with chemotherapy and involved-field radiation therapy (IFRT) in the 1990s represented a major step forward in the treatment of limited-stage HL.26,27 Because of concerns of serious late toxicities (described in detail later), there is universal agreement that the dose and field of radiation should be minimized in the treatment of early-stage HL. DeBruin and associates demonstrated a significant reduction in the incidence of second breast cancers in women treated with mediastinal RT compared with mantle RT for HL, and a meta-analysis of HL trials by Franklin and associates showed a marked decrease in second cancers in patients receiving IFRT compared with EFRT.28,29 However, because mediastinal nodes are involved in the majority of patients with HL, most patients undergoing RT for HL will continue to have at least modest exposure to the heart, lungs, and breasts regardless of efforts to administer only involved-field or even involved-nodal RT. Unfortunately, efforts to completely eliminate radiation for the majority of patients with early-stage HL continue to meet with significant resistance. The 2012 publication of a randomized trial of an ABVD regimen (doxorubicin, bleomycin, vinblastine, dacarbazine) alone versus radiation-based therapy in limited-stage HL showing for the first time a survival advantage in the chemotherapy-alone arm, despite an inferior progression-free survival (PFS), has rekindled the debate.25 Current guidelines and recommendations continue to support either chemotherapy alone or combined-modality therapy in the treatment of nonbulky early-stage HL.30
The specifics of combined-modality therapy for early-stage HL are generally based on the results of two randomized trials from the German Hodgkin Study Group, HD10 and HD11. For patients with the most favorable presentation, two cycles of ABVD and 20 Gy of IFRT resulted in a 5-year freedom from treatment failure (FFTF) of 91% and overall survival rate (OS) of 97%.31 There was no advantage to four cycles of ABVD or 30 Gy of RT in this subset of patients. Importantly, and often misunderstood in the general oncology community, patients eligible for this study (HD10) included only those with disease limited to no more than two sites, no extranodal or bulky disease, and ESR less than 50 if no B symptoms and less than 30 in the presence of B symptoms. Patients with early-stage disease not meeting the eligibility criteria for HD10 were treated on the HD11 trial comparing four cycles of ABVD to four cycles of standard-dose BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone) followed by 20 or 30 Gy IFRT.32 The two chemotherapy regimens yielded equivalent results with less toxicity in the ABVD arm, and the outcomes were superior for those receiving 30 Gy compared with 20 Gy, resulting in the recommendation of four cycles of ABVD plus 30 Gy RT in this less favorable subset. At 5 years, FFTF was 85% and OS was 95%.
Advocates of chemotherapy alone for nonbulky early-stage disease acknowledge the increased failure rate of 5% to 8% when RT is eliminated.25,33 However, with a follow-up duration of 4 to 5 years, no trial comparing chemotherapy alone to combined-modality therapy in early-stage disease has shown a survival advantage for combined modality therapy, likely because second-line strategies, often including RT, result in durable disease control in nearly half of patients with relapsed disease. As alluded to earlier, the National Cancer Institute of Canada (NCIC) study (HD.6) comparing ABVD alone with subtotal nodal irradiation (STNI), with or without ABVD, in 405 patients with nonbulky stage I to IIA HL showed a 12-year PFS of 87% versus 92% (P = .05) in favor of the RT arm of the study but remarkably a 12-year OS of 94% for ABVD alone compared with 87% for those receiving STNI (P = .04). Importantly, 12-year follow-up does not yet reflect the dramatic increase in cardiovascular disease and second malignancies starting 15 to 20 years after primary therapy with RT (Fig. 105-3).34 Critics condemn this study, owing to the use of EFRT in the control arm, arguing that the OS of the RT arm does not reflect modern RT techniques and fields. Regardless of this criticism, the 12-year event-free survival (EFS) of 87% in the ABVD-alone arm of the study compares favorably with the 5- to 8-year EFS rates of 86% to 87% reported in the HD10 and HD11 trials, respectively, of ABVD in combination with limited dose and field RT. In an effort to more accurately compare the results of these trials because of differences in eligibility, staging, end points, and follow-up duration, investigators performed an individual patient-data comparison of HD10 and HD11 and the ABVD-alone arm of HD.6.35 Patients were included if eligible for one of the combined-modality studies as well as the HD.6 study. In 406 patients, the authors concluded that time to progression (hazard ratio [HR], 0.44, 95% confidence interval [CI], 0.24 to 0.78) and PFS (HR, 0.71; CI, 0.42 to 1.18) but not OS trended to being superior in patients treated with combined-modality therapy (HD10/HD11) versus ABVD alone (HD.6). There were no differences in outcomes for patients achieving a complete response (CR) by CT after two cycles of ABVD. IFRT appeared most beneficial to patients who did not achieve an early CR, which provides support for the role of interim response assessment.
The optimum number of cycles of chemotherapy is not defined for patients receiving chemotherapy alone. In the NCIC HD.6 study, 69/196 (35%) of patients treated with ABVD alone achieved CR by CT criteria after two cycles of ABVD.25 Of these 69 CR patients, 57 received a total of four cycles of ABVD (per protocol) and 12 received six cycles (physician decision) with a 5-year freedom from disease progression of 95%, compared with 81% for the 113 patients not in radiographic CR after two cycles (Fig. 105-4). In the small subset of patients with a CR by CT after two cycles of ABVD, a total of four cycles appears adequate. The German Hodgkin Study Group recently adopted the “2 + 2 regimen” for patients with early unfavorable HL based on the results of the HD14 trial.36 Patients were randomized to four cycles of ABVD (ABVD arm) or two cycles of escalated (esc) BEACOPP followed by two cycles of ABVD (2 + 2 arm). Patients in both arms of the study received 30 Gy IFRT after chemotherapy. The 5-year FFTF was superior in the 2 + 2 arm (94.8%) compared with the ABVD arm (87.7%) (P < .001), but there was no difference in OS between the two arms. Grade 3 and 4 toxicities were significantly more frequent in the 2 + 2 arm compared with the ABVD arm. In univariate and multivariate analyses, only large mediastinal mass and elevated ESR were predictors for PFS events.
Based on compelling data from Gallamini and Hutchings showing a 2-year PFS of 12% for HL patients with a positive PET scan after two cycles of ABVD compared with 95% for patients with a negative interim PET, the use of interim PET has become widespread in standard practice and is the focus of most current clinical trials in both early and advanced HL.37 Current studies are examining the use of early metabolic CR by PET/CT to determine the number of cycles of chemotherapy, the intensity of the chemotherapy regimen, and which patients should receive RT.
It is unclear whether the dramatic predictive value of a positive interim PET applies to patients with early-stage HL. In a small study of 57 patients with stage I or II HL, those patients with a positive PET after two cycles of ABVD still had an approximately 75% chance of durable remission with additional ABVD chemotherapy and consolidative IFRT.38,39 Investigators at the Dana Farber Cancer Institute also reported outcomes for 73 patients with HL, 86% with limited stage, referred for consolidative RT, including 20 of 43 patients with a positive interim PET scan and 13 of 73 patients with a positive PET scan at completion of chemotherapy. After IFRT, the 2-year failure-free survival (FFS) was 85% for patients with a positive interim PET and 69% versus 95% for patients with a positive versus negative PET, respectively, at the completion of all chemotherapy.40
Whether a PET-directed approach will allow us to safely minimize chemotherapy cycles and avoid RT is the subject of several ongoing large randomized trials. As these trials complete accrual and results mature, we hope to have a definitive answer as to whether three to four cycles of ABVD alone will be sufficient therapy for the majority of patients with early-stage HL and whether ABVD + RT or alternatively escBEACOPP + RT will be sufficient therapy for patients with less than a complete metabolic response after two cycles of ABVD. Preliminary results of the first completed randomized trial evaluating interim PET-directed therapy in early-stage HL, the UK RAPID trial, provide the earliest support for this concept.41 Eligibility for the RAPID trial included nonbulky HL without regard to other risk factors. All patients received three cycles of ABVD and were restaged with PET/CT. Patients with a negative PET, as defined by a score of 1 or 2 on the London Deauville visual analog scale, were then randomized to receive IFRT versus observation.42 Patients with a positive interim PET went on to receive one additional cycle of ABVD followed by IFRT. In this study, 74.6% of patients had a negative PET scan after three cycles of ABVD. With a median follow-up of 45.7 months, 91.4% of interim PET-negative patients (384/420) and 86.9% of interim PET-positive patients (126/145) are alive and progression free. The 3-year PFS and OS rates were 93.8% and 97%, respectively, in the 209 PET-negative patients randomized to IFRT and 90.7% and 99.5%, respectively, in the 211 PET-negative patients randomized to no further treatment. The PFS survival risk difference of −2.9% [95% CI, −10.5, 1.4%] marginally exceeded the acceptable noninferiority margin of −7%. These very encouraging results support the use of chemotherapy alone in the 75% of patients with early negative interim PET/CT.
Early-Stage Bulky Hodgkin Lymphoma
Despite a paucity of data, the standard treatment for patients with early-stage bulky HL is combined-modality therapy.27,34 The early-stage bulky presentation of HL is common in young women, the population most at risk for late effects of RT, primarily breast cancer. No randomized trials in early-stage bulky HL have specifically addressed the role of RT. The recent intergroup trial E2496 enrolled 854 patients, including 268 with stage I/II bulky disease.43 The patients were treated with ABVD plus modified IFRT to 36 Gy for patients with bulky mediastinal disease or with the Stanford V regimen plus modified IFRT to 36 Gy for sites larger than 5 cm in maximum transverse dimension or in the presence of macroscopic splenic disease. In the patients with bulky disease, the 5-year FFS and OS were 82% and 94%, respectively, with no difference between the study arms. This subgroup analysis provides the “standard” against which trials investigating a reduction or elimination of RT should be compared to determine if the benefits of RT in early-stage bulky disease outweigh the significant risks. The use of fluorodeoxyglucose (FDG)-labeled PET at the end of treatment to guide consolidative radiation was investigated in patients with residual abnormalities on CT.44 Despite omission of RT in patients with a negative end of treatment PET scan, there was no difference in the 3-year time to progression between the patients with bulky disease and those without bulky disease (86% vs. 91%, P = .71). An Italian study of RT versus observation in 160 patients with “bulky” (≥5 cm) disease and a negative end of treatment PET revealed that 14% of patients who did not receive radiation experienced relapse at a median follow-up of 40 months compared with 4% of patients who received RT.45 Until randomized trials are completed, the standard of care for bulky HL remains combined-modality therapy in most centers.
Advanced-Stage Hodgkin Lymphoma
The majority of patients with stage III/IV HL receive six cycles of ABVD chemotherapy, the standard treatment for more than 2 decades. The Hasenclever International Prognostic Score (IPS) is used to predict prognosis of patients with advanced-stage HL treated with ABVD (see Table 105-2).46 In the original publication, the 5-year FFP rates were 70% for those with none to three risk factors compared with 47% for those with four or more high-risk features. A more recent publication examining the outcomes of patients with advanced-stage HL treated in British Columbia between 1980 and 2010 according to the IPS (Table 105-3) showed more favorable 5-year PFS rates of 81% for patients at lower risk (IPS 0 to 3) and 65% for those at high risk (IPS 4 to 7).47 Several reasons are hypothesized for superior PFS rates, including improved diagnostic accuracy resulting in elimination of patients with non-Hodgkin lymphoma, better dose-intensity preservation with current recommendations to treat with full-dose ABVD every 2 weeks regardless of day of treatment neutrophil count, more precise staging with PET/CT resulting in upstaging of perhaps more favorable advanced-stage disease, and anthracycline-based therapy in all patients compared with only 80% of patients in the original series.49–49 The routine use of autologous stem cell transplant (ASCT) in relapsed HL likely accounts for the improvement in OS. Although the IPS continues to delineate outcomes in HL, the narrower range of outcomes makes it more difficult to justify changes in clinical management based on the initial IPS.
Table 105-3
Five-Year Freedom from Progression (FFP) and Overall Survival (OS) by International Prognostic Score (IPS)
| IPS Score | 5-Year FFP | 5-Year OS | ||
| Moccia47 | Hasenclever46 | Moccia47 | Hasenclever46 | |
| 0 | 88 ± 5 | 84 ± 4 | 98 ± 2 | 89 ± 2 |
| 1 | 84 ± 3 | 77 ± 3 | 97 ± 1 | 90 ± 2 |
| 2 | 80 ± 3 | 67 ± 2 | 91 ± 2 | 81 ± 2 |
| 3 | 74 ± 3 | 60 ± 3 | 88 ± 3 | 78 ± 3 |
| 4 | 67 ± 5 | 51 ± 4 | 85 ± 4 | 61 ± 4 |
| ≥5 | 62 ± 7 | 42 ± 5 | 67 ± 7 | 56 ± 5 |
The German Hodgkin Study Group continues to advocate the use of escBEACOPP as initial therapy in all patients with advanced stage HL. Initially reported in 2003, a large randomized trial comparing eight cycles of escBEACOPP to COPP (cyclophosphamide, vincristine, procarbazine, prednisone)/ABVD showed that escBEACOPP resulted in better tumor control and OS.50 Ten-year follow-up of the trial showed FFTF and OS rates of 64% and 75% for COPP/ABVD compared with 82% and 86% for escBEACOPP (P < .0001).51 There was no significant difference in FFTF or OS for patients 60 to 65 years, and escBEACOPP is not recommended in patients 60 years of age or older secondary to unacceptable acute toxicities. Despite the improved FFTF, and perhaps even a survival advantage, physicians outside Germany hesitate to recommend escBEACOPP because of toxicity concerns. Acute hematologic toxicity is high with escBEACOPP, with grade 3 to 4 infections occurring in 22% of patients. Acute myeloid leukemia (AML) developed in 3% of patients on the escBEACOPP arm of the trial. EscBEACOPP results in azoospermia and infertility in the majority of male patients and induces premature menopause in 40% of women younger than age 30 years and 70% of women 30 years of age or older.54–54 An analysis of female survivors treated on the German Hodgkin Study Group HD14 trial revealed a reduced ovarian reserve (as determined by hormone levels) in patients treated with two cycles of escBEACOPP + two cycles of ABVD compared with four cycles of ABVD.55 However, the authors reported that this “2 + 2 regimen” in combination with gonadotropin-releasing hormone (GnRH) analogs did not compromise fertility, because motherhood rates were equivalent to the German normal population. Conflicting reports regarding the use of oral contraceptives and GnRH analogs to protect the ovarian follicle pool have been published.56 A randomized trial of six versus eight cycles of escBEACOPP in advanced-stage HL showed a more favorable toxicity profile and improved 5-year OS with six cycles.57 Secondary AML/MDS (myelodysplastic syndrome) occurred in 0.3% of those receiving six cycles compared with 2.7% for eight cycles, and acute treatment-related deaths were 0.8% with six cycles versus 2.1% for eight cycles.
In three other randomized trials of escBEACOPP versus ABVD, escBEACOPP was not associated with a survival advantage, despite a lower rate of relapse,58,59 All three trials, in an effort to reduce both acute and long-term toxicity, administered a slightly less intensive BEACOPP regimen than the German Hodgkin Study Group with four cycles of escBEACOPP and two cycles of standard BEACOPP in two trials and four cycles of escBEACOPP and four cycles of standard BEACOPP in one study. The lack of a survival difference in these trials is likely due to the better prognosis after second-line therapy for patients initially treated with ABVD compared with escBEACOPP. In one study, 33% (15/45) of the patients who failed ABVD remained in continuous CR after salvage therapy, compared with only 15% (3/20) of the BEACOPP failures at a median follow-up time of 61 months.60
