Cancer Management Chapter 26: Hodgkin lymphoma

March 12, 2010

Article

In 2009 approximately 8,510 new cases of Hodgkin lymphoma (HL) will be diagnosed in the United States. Over the past 4 decades, advances in radiation therapy and the advent of combination chemotherapy have tripled the cure rate of patients with HL. In 2009, more than 80% of all newly diagnosed patients can expect a normal, disease-free life span.

Epidemiology

Gender
The male-to-female ratio of HL is 1.3:1.0.

Age
The age-specific incidence of the disease is bimodal, with the greatest peak in the third decade of life and a second, smaller peak after the age of 50 years.

Race
HL occurs less commonly in African-Americans (2.3 cases per 100,000 persons) than in Caucasians (3.0 per 100,000 persons).

Geography
The age-specific incidence of HL differs markedly in various countries. In Japan, the overall incidence is low, and the early peak is absent. In some developing countries, there is a downward shift of the first peak into childhood.

Studies of young adult HL developing in one of a pair of monozygotic twins suggest a common T helper (Th) cell type 2 (Th2) cytokine profile in each twin with elevated interleukin (IL) -6 levels and decreased IL-12 levels when compared with controls. This inherited cytokine phenotype may increase the likelihood of HL development

(Cozen W et al: Blood 103:3216–3221, 2004; Cozen W et al: Blood 111:3377–3382, 2008)

. Environmental factors also may play a role. The hygiene hypothesis explains a propensity for atopy with an associated lack of a healthy Th cell type 1 response to antigenic challenge noted among patients having decreased fecal-oral exposure during early childhood. Monozygotic twins developing HL showed less fecal-oral exposure and more atopy than did unaffected twins of the same pairs

(Cozen W et al: Blood 112(11): abstract 1457, 2008)

. Thus, genetic and environmental factors may contribute to a Th2 milieu that favors HL development.

Etiology and risk factors

The cause of HL remains unknown, and there are no well-defined risk factors for its development. However, certain associations have been noted that provide clues to possible etiologic factors.

Familial factors
Same-sex siblings of patients with HL have a 10 times higher risk for developing the disease. Patient-child combinations are more common than spouse pairings. Higher risk for HL is associated with few siblings, single-family houses, early birth order, and fewer playmates-all of which decrease exposure to infectious agents at an early age. The monozygotic twin sibling of a patient with HL has a 99 times higher risk of developing HL than a dizygotic twin sibling of a patient with HL. These associations suggest a genetic predisposition and/or a role for an infectious or environmental agent during childhood or early adolescence in the etiology of the disease.

Viruses
Familial aggregation may imply genetic factors, but other epidemiologic findings mentioned previously suggest an abnormal response to an infective agent. Both factors may play a role in the pathogenesis of the disease. The Epstein-Barr virus (EBV) has been implicated in the etiology of HL by both epidemiologic and serologic studies, as well as by the detection of the EBV genome in 20% to 80% of tumor specimens.

There have been no conclusive studies regarding the possible increased frequency of HL in patients with human immunodeficiency virus (HIV) infection. However, HL in HIV-positive patients is associated with an advanced stage and poor therapeutic outcome. (For further discussion of HL in patients with HIV infection, see chapter 24.)

Smoking
The incidence of HL is increased by approximately two-fold in smokers.

Signs and symptoms

HL is a lymph node-based malignancy and commonly presents as an asymptomatic lymphadenopathy that may progress to predictable clinical sites.

Location of lymphadenopathy
More than 80% of patients with HL present with lymphadenopathy above the diaphragm, often involving the anterior mediastinum; the spleen may be involved in about 30% of patients. Less than 10% to 20% of patients present with lymphadenopathy limited to regions below the diaphragm. The commonly involved peripheral lymph nodes are located in the cervical, supraclavicular, and axillary areas; para-aortic pelvic and inguinal areas are involved less frequently. Disseminated lymphadenopathy is rare in patients with HL, as is involvement of Waldeyer's ring and occipital, epitrochlear, posterior mediastinal, and mesenteric sites.

Systemic symptoms
About 30% of patients experience systemic symptoms. They include fever, night sweats, or weight loss (so-called B symptoms) and chronic pruritus. These symptoms occur more frequently in older patients and have a negative impact on prognosis (see section on “Staging and prognosis”).

Extranodal involvement
HL may affect extranodal tissues by direct invasion (contiguity; the so-called E lesion) or by hematogenous dissemination (stage IV disease). The most commonly involved extranodal site is the lungs. Liver, bone marrow, and bone may also be involved.

Diagnosis

The initial diagnosis of HL can only be made by biopsy. Because reactive hyperplastic nodes may be present, multiple biopsies of a suspicious site may be necessary. Needle aspiration is inadequate because the architecture of the lymph node is important for diagnosis and histologic subclassification.

Pathology

Reed-Sternberg cell
In a biopsied lymph node, the R-S cell is the diagnostic tumor cell that must be identified within the appropriate cellular milieu of lymphocytes, eosinophils, and histiocytes. HL is a unique malignancy pathologically in that the tumor cells constitute a minority of the cell population, whereas normal inflammatory cells are the major cell component. As a result, it may be difficult to identify R-S cells in some specimens. Also, other lymphoproliferations may have cells resembling R-S cells.

The R-S cell is characterized by its large size and classic binucleated structure with large eosinophilic nucleoli. Two antigenic markers are thought to provide diagnostic information: CD30 (Ber-H2) and CD15 (Leu-M1). These markers are present on R-S cells and their variants but not on background inflammatory cells. It is also important to obtain a stain for CD20, since it may be positive in the minority of patients with classic HL (nodular sclerosis or mixed cellularity). The prognostic significance of CD20-positive R-S cells in classic HL is controversial.

Studies have confirmed the B-cell origin of the R-S cell. Single-cell polymerase chain reaction (PCR) analysis of classic R-S cells shows a follicular center B-cell origin for these cells with clonally rearranged but crippled V heavy-chain genes, presumably leading to inhibition of apoptosis. Also, high levels of the nuclear transcription factor-kappa-B (NF-kB) have been found in R-S cells; these high NF-kB levels may play a role in pathogenesis by interfering with apoptosis. A molecular link between R-S cells and tumor cells of mediastinal diffuse large B-cell lymphoma has been recently found in gene-profiling studies.

Histologic subtypes
According to the Rye classification (based on the number and appearance of R-S cells, as well as the background cellular milieu), there are four histologic subtypes of HL.

Nodular sclerosis
As the most common subtype, nodular sclerosis is typically seen in young adults (more commonly in females) who have early-stage supradiaphragmatic presentations. Its distinct features are the presence of (1) broad birefringent bands of collagen that divide the lymphoid tissue into macroscopic nodules and (2) an R-S cell variant, the lacunar cell.

Mixed cellularity
This is the second most common histology. It is more often diagnosed in males, who usually present with generalized lymphadenopathy or extranodal disease and with associated systemic symptoms. R-S cells are frequently identified; bands of collagen are absent, although fine reticular fibrosis may be present; and the cellular background includes lymphocytes, eosinophils, neutrophils, and histiocytes.

Lymphocyte-predominant HL
This is an infrequent form of HL in which few R-S cells or their variants may be identified. The cellular background consists primarily of lymphocytes in a nodular or sometimes diffuse pattern. The R-S variants express a B-cell phenotype (CD20-positive, CD15-negative). B-cell clonality has also been demonstrated by PCR of the immunoglobulin heavy-chain genes in single R-S variant cells in biopsy material from patients with lymphocyte-predominant HL.

This finding has led investigators to propose that lymphocyte-predominant HL is a B-cell malignancy with a mature B-cell phenotype, distinct from the other three histologic types of HL. Lymphocyte-predominant HL is often clinically localized, is usually treated effectively with irradiation alone, and may relapse late (a clinical feature reminiscent of low-grade lymphoma). The 15-year disease-specific survival is excellent (> 90%).

The WHO classification recognizes a new subtype of lymphocyte-rich classic HL that has morphologic similarity to nodular lymphocyte-predominant HL. However, the R-S cells have a classic morphology and phenotype (CD30-positive, CD15-positive, CD20-negative), and the surrounding lymphocytes are reactive T cells. This disease subtype does not show a tendency toward late relapse and should be managed like other classic HL histologies.

Lymphocyte depletion
This is a rare diagnosis, particularly since the advent of antigen marker studies, which led to the recognition that many such cases represented T-cell non-HLs (NHLs). R-S cells are numerous, the cellular background is sparse, and there may be diffuse fibrosis and necrosis. Patients usually have advanced-stage disease, extranodal involvement, an aggressive clinical course, and a poor prognosis.

Staging and prognosis

Precise definition of the extent of nodal and extranodal involvement with HL according to a standard staging classification system is critical for selection of the proper treatment strategy.

Staging system
The staging system is detailed in Table 1, and the anatomic regions that provide the basis for the

staging classification are illustrated in Figure 1. The assignment of stage is based on:

• the number of involved sites
• whether lymph nodes are involved on both sides of the diaphragm and whether this involvement is bulky (particularly in the mediastinum)
• whether there is contiguous extranodal involvement (E sites) or disseminated extranodal disease
• whether typical systemic symptoms (B symptoms) are present.

In defining the disease stage, it is important to note how the information was obtained, since this fact reflects on remaining uncertainties in the evaluation for extent of disease. Clinical staging refers to information that has been obtained by initial biopsy, history, physical examination, and laboratory and radiographic studies only. A pathologic stage is determined by more extensive surgical assessment of potentially involved sites, eg, by surgical staging laparotomy and splenectomy.

Also, various designations relating to the presence or absence of B symptoms or bulky disease (see Table 1) can be applied to any disease stage. For example, a patient with no B symptoms but with a bulky mediastinal mass and involvement of the cervical lymph nodes would be defined as having CS IIAX disease. A patient with axillary disease and fever who underwent a staging laparotomy that revealed involvement of the para-aortic lymph nodes and spleen would be staged as PS III2B.

Most recent studies in stage I/II disease distinguish between favorable and unfavorable early-stage disease, according to the EORTC definitions outlined in Table 2.

Clinical staging evaluation

Disease-associated symptoms
As mentioned previously, disease-associated symptoms may occur in up to one-third of patients. They may include B symptoms, pruritus, and, less commonly, pain in involved regions after ingestion of alcohol. In each anatomic stage, the presence of B symptoms is an adverse prognostic indicator and may strongly affect treatment choices. B symptoms are carefully defined in the staging system. Unexplained fever should be > 38°C and recurrent during the previous month, night sweats should be drenching and recurrent, and unexplained weight loss should be significant only if > 10% of body weight has been lost within the preceding 6 months. Although pruritus is no longer considered to be a B symptom, the presence of generalized itching may be considered to be an adverse prognostic symptom.

Certain combinations of B symptoms are more prognostically significant than others. For example,

the combination of fever and weight loss has a worse prognosis than do night sweats alone.

Physical examination
The physician should carefully determine the location and size of all palpable lymph nodes. Inspection of Waldeyer's ring, detection of splenomegaly or hepatomegaly, and evaluation of cardiac and respiratory status are important.

Laboratory studies
These should include a complete blood cell count with white blood cell (WBC) differential and platelet count, the erythrocyte sedimentation rate (ESR), tests for liver and renal function, and assays for serum alkaline phosphatase and lactate dehydrogenase (LDH). A moderate to marked leukemoid reaction and thrombocytosis are common, particularly in symptomatic patients, and usually disappear with treatment.

ESR studies may provide helpful prognostic information. At some centers, treatment programs for patients with early-stage disease are influenced by the degree of ESR elevation. In addition, changes in the ESR following therapy may correlate with response and relapse.

Abnormalities of liver function studies should prompt further evaluation of that organ, with imaging and possible biopsy.

Alkaline phosphatase An elevated alkaline phosphatase level may be a nonspecific marker, but it may also indicate bone involvement that should be appropriately evaluated by a radionuclide bone scan and directed skeletal radiographs.

Imaging studies
Radiologic studies should include a chest x-ray and computed tomography (CT) scan of the chest, abdomen, and pelvis with IV contrast. Positron emission tomography (PET) scan will provide important information on the extent of disease and a baseline for evaluation of response to treatment. Radionuclide bone scan, magnetic resonance imaging (MRI) of the chest or abdomen, and CT scan of the neck are contributory only under special circumstances.

Evaluation for supradiaphragmatic disease The thoracic CT scan details the status of intrathoracic lymph node groups, the lung parenchyma, pericardium, pleura, and chest wall. Since the chest CT scan may remain abnormal for a long time after the completion of therapy, the PET scan is used to evaluate pretreatment involvement and response to therapy.

Evaluation of the abdomen and pelvis The CT scan and PET scan are basic imaging studies for evaluating the abdomen. They are recommended as part of initial staging and are critical in the new International Working Group Criteria for assessing response after treatment. A change from a positive pretreatment PET/CT scan to negative following completion of therapy is considered to be the criterion for complete response even if a residual mass is found on CT scan.

Bone marrow biopsy
Bone marrow involvement is relatively uncommon with HL, but because of the impact of a positive biopsy on further staging and treatment, unilateral bone marrow biopsy should be part of the staging process of patients with stage IIB disease or higher.

Lymphangiography and staging laparotomy
These two old methods of staging have been replaced by modern imaging techniques using high-resolution CT scanning and 18F-fluorodeoxyglucose (FDG)-PET.

Treatment

HL is sensitive to radiation and many chemotherapeutic drugs, and, in most stages, there is more than one effective treatment option. Disease stage is the most important determinant of treatment options and outcome. All patients, regardless of stage, can and should be treated with curative intent.

Treatment of stage I/II disease
A common treatment choice for favorable and unfavorable early-stage HL is brief chemotherapy

followed by involved-field radiotherapy (IFRT). Most of the experience that yielded excellent treatment results with low toxicity was with ABVD (Adriamycin [doxorubicin], bleomycin, vinblastine, and dacarbazine) for 4 cycles and IFRT of 30 to 36 Gy. Table 3 summarizes data from randomized studies that reported on the combination of short chemotherapy (4 or even only 2 cycles) followed by IFRT. The three top randomized studies have also indicated that adding extended-field radiotherapy (EFRT) to chemotherapy is not necessary and the small involved field is adequate.

The most recent (and, as yet, not fully mature) excellent results are with shortening the duration of chemotherapy to only 2 cycles of ABVD in favorable patients and reducing the IFRT dose to 20 Gy (Table 3). The addition of consolidative IFRT to chemotherapy for patients with bulky disease, which usually appears in the mediastinum, remains standard treatment. However, the role of PET scans in assessing response during or after chemotherapy and the possibility of omitting radiotherapy in patients with negative PET scans are both controversial and the subject of ongoing and planned clinical trials.

Subtotal lymphoid irradiation
This therapeutic modality (ie, treatment of the mantle and para-aortic fields only) remains an adequate alternative treatment of clinically or pathologically staged favorable (nonbulky and without B symptoms) early-stage HL (stage I/II). Yet this option is no longer the treatment of choice due to the risk of second tumors and (to a lesser degree) coronary artery disease in long-term survivors of extensive radiotherapy alone as practiced in the past. In classic (non–lymphocyte-predominant) HL, subtotal lymphoid irradiation is adequate for patients who are not candidates for a chemotherapy-containing strategy.

In patients who underwent pathologic staging (laparotomy) and were treated with primary irradiation alone, several large series reported a 15- to 20-year survival rate of nearly 90% and a relapse-free survival rate of 75% to 80%. Most relapses (75%) occurred within 3 years after the completion of therapy; late relapses were uncommon. More than half of the patients who relapsed after radiotherapy alone were still curable with standard chemotherapy.

Two recently published, large, randomized trials, the EORTC H7-F trial and the GHSG HD7 trial, further established significant decreases in relapse rates at 7 and 10 years, respectively, for combined modality chemotherapy and IFRT as compared with EFRT alone. However, neither study showed a difference in survival between the two arms.

Canadian and European studies have reported excellent overall survival results in patients selected for radiotherapy on the basis of clinical prognostic factors alone. Thus, irradiation alone can be safely offered to clinically staged patients with favorable prognostic factors who are not candidates for combined-modality treatment.

Chemotherapy alone
In two prospective, randomized studies, radiotherapy alone was as effective as or superior to MOPP (mechlorethamine [Mustargen], Oncovin [vincristine], procarbazine [Matulane], and prednisone) chemotherapy in improving the survival of patients with early-stage disease. Although the relapse rate after MOPP was similar to that after radiotherapy, conventional-dose salvage chemotherapy used after failure of chemotherapy produces poor results, which translates into an inferior overall survival.

More recently, five prospective randomized studies compared chemotherapy alone with chemotherapy followed by IFRT or regional radiotherapy in patients with early-stage HL. The Children's Cancer Group tested the role of radiation therapy in young patients (< 21 years old) who attained a complete response with risk-adapted chemotherapy (mostly COPP [cyclophosphamide, Oncovin (vincristine), procarbazine, and prednisone)/ABV [Adriamycin (doxorubicin), bleomycin, and vinblastine], 4 to 6 cycles). They enrolled 829 patients into the study (68% had early-stage disease); 501 patients who achieved a complete response were then randomized to receive either low-dose (21 Gy) IFRT or no further treatment. The accrual was stopped earlier than planned because of a significantly higher number of relapses on the no-radiotherapy arm. The 3-year event-free survival rate with an intent-to-treat analysis was 92% for patients randomized to receive radiotherapy and 87% for those randomized to receive no further treatment (P = .057).

The EORTC/Group d'Etude des Lymphomes de l'Adulte (GELA) conducted a large randomized trial in patients with favorable early-stage classic HL. All patients received 6 cycles of EBVP (epirubicin, bleomycin, vinblastine, and prednisone). Only patients who achieved a complete response were randomized to receive either IFRT of 36 Gy, IFRT of 20 Gy, or no irradiation. After an interim analysis, the EORTC/GELA groups closed the entry to the no-radiotherapy arm because of the excessive number of relapses. It should be noted that in previous EORTC studies, EBVP with IFRT was found to be inferior to MOPP/ABV with IFRT hybrid in patients with unfavorable disease but provided excellent results when combined with radiotherapy and was compared with radiotherapy alone in patients with favorable disease.

The NCIC and ECOG included 405 patients with nonbulky stage I/II disease. They were randomized to receive either “standard therapy,” namely, subtotal nodal irradiation (STNI) for favorable patients, and ABVD (2 cycles) followed by STNI for unfavorable (B symptoms, elevated ESR, ≥ 3 sites, age ≥ 40, mixed cellularity histology) patients or experimental therapy consisting of 6 cycles or 4 cycles (if complete response was attained after 2 cycles) of ABVD and no radiotherapy. At a median follow-up of 4.2 years, progression-free survival with ABVD alone was significantly inferior (P = .006; hazard ratio [HR] = 2.6; 5-year estimates of disease progression-free survival, 87% vs 93%). At 5 years, no event-free or overall survival difference has been detected. The study was planned for 12 years' analysis of survival.

The MSKCC trial included 152 patients with nonbulky, early-stage HL. Patients were randomized up front to receive either ABVD for 6 cycles alone or ABVD for 6 cycles followed by radiotherapy. At 60 months, the duration of complete response and freedom from disease progression for ABVD and radiotherapy versus ABVD alone were 91% versus 87% (P = .61) and 86% versus 81% (P = .61), respectively. Overall survival was 97% with ABVD and radiotherapy vs 90% with ABVD alone (P = .08). Although the differences between the outcome of the two treatment groups were not statistically significant, the study was not powered to detect differences between the treatment strategies that were smaller than 20%, due to the small number of patients and events.

In a prospective, randomized study reported from India, patients with HL who achieved a complete response after ABVD were randomized to receive either IFRT or no further therapy. The 8-year event-free and overall survival rates were significantly better for the patients who received consolidation with IFRT than for those who received ABVD alone. Subset analysis indicated that the benefit from added IFRT was more prominent in advanced-stage than in early-stage disease.

The NCCN guidelines recommend combined-modality therapy as the treatment of choice for favorable or unfavorable classic HL. Opinions differ as to the use of chemotherapy alone, and it remains controversial.

However, because of concern about the late morbidity and mortality related to radiotherapy including second malignancies and cardiovascular complications, some medical oncologists favor the use of chemotherapy only for patients with non-bulky, stages I and II HL. The treatment regimen employed is usually 6 cycles of ABVD or 4–6 cycles of ABVD (complete response + 2 additional cycles).

Technical aspects of radiation therapy

IFRT and involved-node radiotherapy (INRT)
In a combined-modality setting, irradiation of the involved lymph node chain, with or without adjacent sites, and tailoring of the field borders to the postchemotherapy tumor volume (in critical areas such as the mediastinum) are recommended. IFRT is the most appropriate irradiation approach after chemotherapy. IFRT or regional radiotherapy alone is used for lymphocyte-predominant HL. Recommendations for IFRT for HL are detailed in an article by Yahalom and Mauch. Recently, further reduction in IFRT has been introduced and implemented in EORTC/GELA studies. The INRT field is limited to the prechemotherapy lymph node(s) volume, and, in the mediastinum, it also accounts for the reduction of mass after chemotherapy.

The transition into INRT fields or reduced IFRT markedly reduces the exposure of normal structures in most cases; the safety of field reduction has recently been reported (see side bar).

The lymphoma team from British Columbia Cancer Agency analyzed the implications of further reducing the radiation field following a short course of chemotherapy. In all, 325 patients with limited-stage HL were treated with only 2 cycles of ABVD. During the early period (1989–2005), all patients also received EFRT. From 1996–2001, ABVDX2 was followed by IFRT. More recently (2001–present), the radiation field was reduced by 50% to a size similar to that of the increasingly popular INRT. The three groups had similar prognostic characteristics, and the median follow-up was 80 months. The very small relapse rate (3%–5%) was similar among all groups, and no marginal misses occurred after INRT. The 10-year progression-free and overall survival rates were 95% and 90%, respectively, for all patients, with no difference found between the various radiation fields. Although retrospective, the results established the safety of the very limited field radiotherapeutic approach

(Campbell BA et al: J Clin Oncol 26:5170–5174, 2008)

Extended radiation fields
Successful therapy with irradiation alone in classic HL requires treatment of all clinically involved lymph nodes and all nodal and extranodal regions at risk for subclinical involvement (Figure 2). The HL radiation fields were designed to conform to the philosophy of treating regions beyond the immediately involved area while accounting for normal tissue tolerance and the technical constraints of field size. The extended radiation fields are inappropriate when radiation is administered as consolidation following chemotherapy and, thus, they are rarely used today.

Dose considerations
When irradiation alone is used to treat HL, the standard total dose to each field is 3,600 cGy, delivered in daily fractions of 180 cGy over 4 weeks. In addition, clinically involved areas are given a boost of 360 to 540 cGy in 2 to 3 fractions to bring the total dose to these areas up to 3,960 to 4,140 cGy. Patients who receive irradiation as consolidation after chemotherapy receive a total dose of 2,000 to 3,600 cGy in 150- to 180-cGy fractions. We normally use opposed anterior and posterior fields that are evenly weighted and treat both fields daily. Three-dimensional conformal radiotherapy and intensity-modulated radiation therapy (IMRT) are employed for selected cases.

Recent studies suggest that a reduction in the size of radiation fields and doses delivered may decrease the risk of breast cancer.

Side effects and complications of radiotherapy
Side effects of radiotherapy depend on the irradiated volume, dose administered, and technique employed. They are also influenced by the extent and type of prior chemotherapy, if any, and by the

patient's age.

Acute effects
The potential acute side effects of involved fields in the upper body include mouth dryness, change in taste, pharyngitis, nausea, dry cough, dermatitis, and fatigue. These side effects are usually mild and transient.

The main potential side effects of subdiaphragmatic irradiation are loss of appetite, nausea, and increased bowel movements. These reactions are usually mild and can be minimized with standard antiemetic medications.

Irradiation of more than one field, particularly after chemotherapy, can cause myelosuppression, which may necessitate treatment delays.

Delayed effects
Delayed side effects may develop anywhere from several weeks to several years after the completion of radiotherapy.

Lhermitte's sign Approximately 15% of patients who receive the full dose of radiation to the neck may note an electric shock sensation radiating down the backs of both legs when the head is flexed (Lhermitte's sign) 6 weeks to 3 months after mantle-field radiotherapy. Possibly secondary to transient demyelinization of the spinal cord, Lhermitte's sign resolves spontaneously after a few months and is not associated with late or permanent spinal cord damage.

Pneumonitis and pericarditis During the same period, radiation pneumonitis and/or acute pericarditis may occur in < 5% of patients who receive large fields of radiation to the mediastinum; these side effects occur more often in those who have extensive mediastinal disease. Both inflammatory processes have become rare with modern radiation techniques.

Herpes zoster infection Patients with HL, regardless of treatment type, have a propensity to develop herpes zoster infection within 2 years after therapy. Usually, the infection is confined to a single dermatome and is self-limited. If the cutaneous eruption is identified promptly, treatment with systemic acyclovir will limit its duration and intensity.

Subclinical hypothyroidism
Radiotherapy of the neck can induce subclinical hypothyroidism in about one-third of patients. This condition is detected by elevation of thyroid-stimulating hormone. Thyroid replacement with levothyroxine is recommended, even in asymptomatic patients, to prevent overt hypothyroidism and decrease the risk of benign thyroid nodules.

Infertility
Irradiation of the pelvis may have deleterious effects on fertility. In most patients, this problem can be avoided by appropriate gonadal shielding. In females, the ovaries can be moved into a shielded area laterally or inferomedially near the uterine cervix. Irradiation fields that spare the pelvis do not increase the risk of sterility.

Secondary malignancies
Patients with HL who were cured with radiotherapy and/or chemotherapy have an increased risk of secondary solid tumors (most commonly, lung, breast, and stomach cancers, as well as melanoma) and NHL 10 or more years after treatment. Chemotherapy combinations that do not include alkylating agents or a new brief program (such as Stanford V) as well as the smaller involved fields and lower doses are less likely to have the increased risk observed with extended fields and/or MOPP or MOPP-like chemotherapy.

Lung cancer Patients who are smokers should be strongly encouraged to quit the habit because the increase in lung cancer that occurs after irradiation or chemotherapy with alkylating agents has been detected mostly in smokers. Alkylating agents (such as in the MOPP regimen) and radiation therapy were associated with an increased risk of lung cancer in an additive and dose-dependent fashion. These effects were multiplied by tobacco use (see “Suggested Reading”).

Breast cancer The increase in breast cancer risk is inversely related to the patient's age at HL treatment; no increased risk has been found in women irradiated after 30 years of age. The risk of breast cancer is increased with higher radiation breast dose and is reduced in patients who received chemotherapy or ovarian irradiation that induced early menopause.

In most situations, modern IFRT should spare the breast. Breast cancer is curable in its early stages, and early detection has a significant impact on survival. Breast examination should be part of the routine follow-up for women cured of HL, and routine mammography should begin about 8 years after treatment.

Cardiovascular disease
An increased risk of cardiovascular morbidity has been reported among patients who have received mediastinal irradiation. In a retrospective study evaluating the cardiac risks of 450 patients cured of HL with radiotherapy alone or in combination, 42 patients (10%) developed coronary artery disease (CAD) at a median of 9 years after treatment, 30 patients (7%) developed carotid and/or subclavian artery disease at a median of 17 years after treatment, and 25 patients (6%) developed clinically significant valvular dysfunction at a median of 22 years after treatment. The most common valve lesion was aortic stenosis, which occurred in 14 valves. The only treatment-related factor associated with the development of CAD was use of a radiation technique that resulted in a higher total dose to a portion of the heart (relative risk [RR] = 7.8; 95% CI,1.1–53.2; P=.04). No specific treatment-related factor was associated with carotid or subclavian artery disease or valvular dysfunction. Freedom from cardiovascular morbidity was 88% at 15 years and 84% at 20 years.

To reduce this hazard, radiation fields should conform to the involved postchemotherapy volume, and the dose should be reduced to 20 to 30 Gy, if possible. Patients who have received radiation to the mediastinum should be monitored and advised about other established CAD risk factors, such as smoking, hyperlipidemia, hypertension, and poor dietary and exercise habits. Cholesterol levels should be monitored and treated if elevated.

The risk of late mortality from MI is increased with use of supradiaphramatic radiotherapy, especially total nodal radiation therapy. The risk also was higher in patients given various chemotherapy regimens (particularly ABVD) with, and even without, supradiaphragmatic radiotherapy.

Effects on bone and muscle growth
In children, high-dose irradiation will affect bone and muscle growth and may result in deformities. Current treatment programs for pediatric HL are chemotherapy-based; radiotherapy is limited to low doses.

Treatment of stage III/IV disease
Chemotherapy has become curative for many patients with advanced stages of HL. MOPP has been the primary effective combination chemotherapy regimen for advanced-stage disease since the 1960s. Over the past several years, ABVD has been shown to be more effective and less toxic than MOPP, particularly with respect to sterility and secondary leukemia.

Combination chemotherapy regimens
Doxorubicin-containing regimens A doxorubicin-containing regimen, such as ABVD (Table 4), is

the treatment of choice for patients presenting with stage III or IV disease, as demonstrated by a randomized phase III trial undertaken by the CALGB. This trial showed higher complete response rates with ABVD and ABVD/MOPP (82% and 83%, respectively) than with MOPP alone (65%).

One reason for the improved response rate in the groups treated with doxorubicin-containing regimens was the higher percentage of patients who were able to receive ≥ 85% of the expected chemotherapy dose, particularly in the ABVD group. In addition, rates of significant and life-threatening neutropenia were higher in patients treated with the MOPP-containing regimens than in those treated with other regimens.

Subsequent trials compared ABVD, alternating MOPP/ABVD, and a MOPP/ABV hybrid. Alternating MOPP/ABVD and the MOPP/ABV hybrid was found to be equally effective in treating advanced-stage HL. However, an intergroup study that compared ABVD with MOPP/ABV hybrid (without irradiation) was closed early because of concerns of excess treatment-related deaths and second malignancies (mostly acute myelogenous leukemia and lung cancer) in the MOPP/ABV hybrid arm. ABVD and MOPP/ABV hybrid yielded similar 5-year failure and overall survival rates.

Shortened dose-intense regimens Shortened dose-intense regimens have shown promise. For example, the 12-week Stanford V regimen (see Table 4) combined with IFRT produced a 5-year overall survival rate of 96% and a freedom-from-disease-progression rate of 89%. The freedom-from-disease-progression rate was significantly superior among patients with a prognostic score of 0–2, compared with those with a score of 3 and higher (94% vs 75%; P = .0001). Of interest, in 142 patients from Stanford, no secondary leukemia was observed, and 42 pregnancies were reported.

An escalated dose version of BEACOPP (bleomycin, etoposide, Adriamycin [doxorubicin], cyclophosphamide, Oncovin [vincristine], procarbazine, and prednisone) was found to have a statistically significant superior freedom from treatment failure at 5 years compared with standard-dose BEACOPP and alternating monthly COPP and ABVD for patients with advanced stages of HL. Short-term hematologic toxicity was greatest for escalated BEACOPP, and a significant increased risk for secondary acute leukemias was also seen as compared with standard-dose BEACOPP and COPP/ABVD.

In advanced-stage disease, PET scanning during or after chemotherapy is emerging as a powerful predictor of outcome. Patients with advanced-stage disease who have positive PET scans after 2 cycles of ABVD are at a higher risk of relapse following treatment than are patients who have negative PET scans. Risk-adapted clinical trials based on this finding, which also holds true in early-stage HL, currently are ongoing and opening in the United States and Europe. At present, a change in therapy due to PET scan results during treatment cannot be recommended outside of the setting of a clinical trial. The PET scan currently is standard for assessing response after the completion of therapy.

Combined-modality therapy
Although the role of consolidation radiotherapy after induction chemotherapy remains controversial, irradiation is routinely added in patients with advanced-stage disease who present with bulky disease or who remain in uncertain complete remission after chemotherapy. Retrospective studies have demonstrated that adding low-dose radiotherapy to all initial disease sites after chemotherapy-induced complete response decreases the relapse rate by ~25% and significantly improves overall survival.

Interpretation of the impact of irradiation in prospective studies has been controversial. However, a SWOG randomized study of 278 patients with stage III or IV HL suggested that the addition of low-dose irradiation to all sites of initial disease after a complete response to MOP-BAP (mechlorethamine, Oncovin [vincristine], prednisone, bleomycin, Adriamycin [doxorubicin], and procarbazine) chemotherapy improves the duration of remission in patients with advanced-stage disease. An intention-to-treat analysis showed that the advantage of combined-modality therapy was limited to patients with nodular sclerosis. No survival differences were observed.

A meta-analysis demonstrated that the addition of radiotherapy to chemotherapy reduced the rate of relapse but did not show a survival benefit for the combined-modality approach.

The GHSG HD12 Study compared 8 cycles of escalated BEACOPP with 4 cycles of standard BEACOPP in patients with advanced-stage HL. In each arm, there was a second randomization to IFRT or no radiotherapy. The trial did not achieve statistical power to determine the primary endpoint of 5-year freedom from treatment among the four arms, but a subset analysis showed no difference between the chemotherapy arms

(Diehl V et al: Blood 112(11): abstract 1558, 2008)

. A randomized trial confirmed the superior freedom from first relapse with use of 4 escalated and 4 standard cycles of BEACOPP when compared with more standard ABVD (comparable with COPP and ABVD), but freedom from second progression and ultimate overall survival among patients given ASCT after high-dose chemotherapy and radiotherapy were the same among the two arms

(Gianni AM et al: J Clin Oncol 26[15S]:abstract 8506, 2008)

. This outcome, along with a treatment-related mortality of ~3% with use of 4 escalated and 4 standard BEACOPP cycles in these trials, has somewhat dampened enthusiasm in this country for escalated BEACOPP, except for the small subgroup of advanced-stage patients having multiple adverse risk factors.

The EORTC conducted a randomized trial in patients with stages III and IV HL in which those achieving a complete remission with MOPP/ABV hybrid were randomized to receive either low-dose IFRT or no radiotherapy. Of the 739 patients enrolled, 421 achieved a complete remission. The median follow-up was 79 months. There was no statistically significant difference in 5-year event-free or overall survival. Partial responders received low-dose IFRT, and their event-free and overall survival rates were similar to those patients who achieved a complete remission.

Long-term toxicities of combination chemotherapy
The CALGB trial and the intergroup trials mentioned previously (see section on “Combination chemotherapy regimens”) noted differences in the long-term toxicities of various combination chemotherapeutic regimens (Table 5).

Myelodysplasia and acute leukemia
MOPP therapy is known to be related to the development of myelodysplastic syndromes and acute leukemia. These secondary hematologic malignancies begin 2 years following therapy and decline by 10 years, with the maximum risk between 5 and 9 years. Patients with these malignancies have a poor prognosis.

The incidence of secondary leukemia appears to increase with cumulative doses of chemotherapy, age > 40 years when receiving chemotherapy for HL, and splenectomy. It is controversial whether combined-modality therapy increases the risk of leukemia compared with chemotherapy alone.

Cytogenetic studies of secondary leukemias reveal a loss of the long arm of chromosome 5 and/or 7. Less frequently, there is a loss of chromosome 18 or rearrangement of the short arm of chromosome 17. A balanced rearrangement of 11q23 and 2lq22 also has been described with etoposide therapy.

Other malignancies
Other types of cancer also are being observed with increasing frequency after chemotherapy (most regimens included alkylating agents), particularly for lung cancer and NHL. These malignancies have a longer latency period and usually are not observed until 15 years after therapy.

Infertility
The inability to conceive a child after trying for 1 year is another long-term complication seen with combination chemotherapy. At least 80% of males have permanent azoospermia or oligospermia following more than 3 cycles of MOPP chemotherapy; < 10% of men will have recovery of spermatogenesis within 1–7 years following the end of chemotherapy. The risk of infertility with ABVD chemotherapy is significantly lower (~15%–25%) than that with MOPP. All men who desire childbearing potential following therapy should be counseled regarding sperm banking.

In females, there is a 50% rate of primary ovarian failure overall. The risk is 25% to 30% in patients treated at age 25 or younger but increases to 80% to 100% in women older than age 25. Many women who maintain ovarian function during chemotherapy will have premature menopause following therapy.

Female fertility appears to be well preserved following administration of ABVD, which does not contain alkylating agents of the nitrogen mustard type or procarbazine. Male fertility also probably is better preserved with such treatment, although this finding is not as well documented.

Pulmonary complications
Lung problems have been reported with ABVD chemotherapy and are related to bleomycin-induced lung toxicity. In a MSKCC study of 60 patients with early-stage HL receiving ABVD chemotherapy with or without mediastinal irradiation, bleomycin was discontinued in 23% of patients. Following ABVD therapy, there was a significant decline in median forced vital capacity (FVC) and diffusing capacity of the lungs for carbon monoxide. Radiotherapy following ABVD chemotherapy resulted in a further decrease in FVC but did not significantly affect functional status. In a study from the Mayo Clinic, bleomycin pulmonary toxicity (BPT) was observed in 18% of patients. Increasing age, and use of ABVD and granulocyte colony-stimulating factor were associated with development of BPT. Patients with BPT had a 5-year overall survival of only 63% as compared with 90% (P = .001) in patients without BPT. Mortality from BPT was 4.2% in all patients and 24% in those who developed BPT. The omission of bleomycin had no effect on obtaining complete remission or on progression-free or overall survival.

In the CALGB trial, there were 3 fatal pulmonary complications in 238 patients; all 3 patients were older than age 40.

Pulmonary fibrosis has also been described after combined-modality therapy. Pulmonary function testing usually reveals a decreased diffusion capacity and restrictive changes prior to the onset of symptoms.

Cardiomyopathy
Deteriorated myocardial function is a recognized complication of doxorubicin therapy but is not commonly seen in patients receiving ABVD chemotherapy. Patients who are treated with 6 cycles of ABVD chemotherapy receive a total doxorubicin dose of 300 mg/m²; cardiac toxicity is rarely seen in patients who receive a total dose ≤ 400 mg/m².

Management of relapsed disease

Relapse after radiation therapy
Patients with early-stage HL who relapse after initial therapy with irradiation alone have excellent complete remission rates and 50% to 80% long-term survival rates when treated with MOPP or ABVD. The dose regimens used for salvage therapy are the same as those outlined in Table 4.

Relapse after combination chemotherapy
Among patients with advanced-stage HL, 70% to 90% will have a complete response to treatment; however, up to one-third of patients with stage III or IV disease will relapse, usually within the first 3 years after therapy.

Various studies have identified the following poor prognostic factors for response to first-line chemotherapy: B symptoms, age > 45 years, bulky mediastinal disease, extranodal involvement, low hematocrit, high ESR, high levels of CD30, and high levels of serum interleukin-10 (IL-10) and soluble IL-2 receptor.

An International Prognostic Index (IPI) has been devised for advanced HL based on a retrospective analysis of 1,618 patients from 25 centers. In the final model, seven factors were used: albumin < 4 g/dL, hemoglobin < 10.5 g/dL, male gender, stage IV disease, age ≥ 45 years, WBC ≥ 15,000/μL, and lymphocytes < 600/μL (or 8% of the WBC count). The worst prognostic group (7%) had a 5-year overall survival rate of 56% and a failure-free survival rate of 42%.

In a comparison of seven well-known prognostic models for HL applied retrospectively to a population of patients with advanced-stage disease, three were found to be the most predictive of outcome. One was the IPI, and the other two were the MSKCC model (employing age, LDH, hematocrit, inguinal nodal involvement, and mediastinal mass bulk) and the Database on Hodgkin Lymphoma model (employing stage, age, B symptoms, albumin level, and gender). Integration of the three models in a linear model improved their predictive power.

Long-term effects and quality of life of 153 patients with relapsed or refractory HL who survived ≥ 2 years following salvage with high-dose chemotherapy and radiation followed by autologous stem cell rescue were reported by the MSKCC transplantation team. Median follow up was 11.5 years. Of 53 deaths, 33 resulted from HL and 20 from other causes. The most common cause of non-HL death was second malignancy with a relative risk of 6.5 when compared with the general population, but only 2.4 (CI 1.4–4) when compared to HL patients (SEER data). Global quality of life of survivors was comparable with that of the general population, but specific domains indicated some reduced functioning and worsened symptoms

(Goodman KA et al: J Clin Oncol 26:5240–5247, 2008)

High-dose therapy with autologous stem-cell transplantation The preferred salvage method for patients who relapsed after combined-modality therapy or chemotherapy alone or remained refractory to those programs is high-dose chemoradiotherapy with autologous stem-cell transplantation (ASCT).

Two randomized studies (from Great Britain and Germany) demonstrated an event-free survival advantage with the high-dose therapy approach. Although a significant survival advantage was not observed due to the crossover design of the studies, most patients with refractory disease or postchemotherapy relapse are currently managed with high-dose chemoradiation therapy and ASCT.

No standard conditioning regimen has been used in this setting, as patients have had prior treatment with a variety of combinations of chemotherapy and radiation therapy. Most patients who have received bone marrow have been treated with several regimens or have had poorly responsive disease from initial diagnosis, but the complete response rate has ranged from 50% to 80%, with approximately 40% to 80% of responding patients achieving durable remission.

Analysis of prognostic factors in patients receiving high-dose salvage therapy indicated that B symptoms at relapse, extranodal disease, and short (< 1 year) remission or no remission are associated with a poor outcome.

A study from MSKCC reported the results of high-dose chemotherapy with ASCT in 65 patients with relapsed or refractory HL. At a median follow-up of 43 months, overall survival was estimated to be 73%, and event-free survival was estimated to be 58% by intent-to-treat analysis. In a multivariate logistic regression model, there were three adverse prognostic factors: extranodal sites of relapse or refractory disease, complete remission duration of < 1 year or refractory disease, and B symptoms. Patients with no or one adverse factor had an overall survival of 90% and an event-free survival of 83%, those with two adverse factors had an overall survival of 57% and an event-free survival of 27%, and those with three adverse factors had an overall survival of 25% and an event-free survival of 10%. A follow-up study of a risk-adapted approach based on the study previously described suggested that patients with adverse prognostic factors may benefit from further augmentation of high-dose programs, including a “double-transplant” for selected patients.