Pediatric Annals

Non-Hodgkin's Lymphoma

Renee V Gardner, MD; John Graham-Pole, MD

Abstract

The heterogeneous group of disorders known as nonHodgkin's Iymphoma (NHL), fortunately, is uncommonly seen in children. Although its incidence is not exactly defined, it is recognized as being one of the commonest malignancies of childhood, comprising almost 10% of childhood cancers and 60% of lymphomas. The clinical manifestations of non-Hodgkin's Iymphoma contrast strongly with Hodgkin's disease which is, forthe most part, an indolent process, occurring in a somewhat older age group and characterized by contiguous spread of the tumor. The adult and childhood forms of nonHodgkin's Iymphoma also show distinct differences. Thus, the pediatrie variant is characterized by noncontiguous dissemination, a tendency for early !eukemic conversion, mediastinal involvement, and invasion of the meninges.

Non-Hodgkin's Iymphoma is seen predominantly in the five- to 15-year age range, with a median age of eight years; it is more common in males than females by a ratio of 2:1 to 3:1. It is rarely seen under the age of one year.

The recognition of clinical features of this disease that are peculiar to children has led to recent advances in the prognosis and treatment of children with NHL. Patterns of response or of tumor behavior are now being discerned and analyzed, and the knowledge applied. Twenty years ago, the survival rate for children with non-Hodgkin's Iymphoma was under 10%. More aggressive therapeutic approaches have since been used in major centers, with marked improvement in long-term, disease-free survival.

ETIOLOGY

Type C RNA viruses and certain DNA viruses, most notably the herpes viruses, have been shown to have oncogenic potential among non-human mammalian and avian species.1 Presumably, after a period of latency, endogenous viral replication may be stimulated by exogenous exposure, such as to x-rays, chemicals, or infectious agents, leading to the development of lymphoid malignancy. Evidence to support the concept of viral oncogenesis has stemmed from the isolation of viral particles or of reverse transcriptase from tumor cells, and serologìe evidence of co-existent or pre-existent viral infection.

However, the isolation of viral material from human leukemia or Iymphoma cells is not a consistent finding, nor has there been consistent serologie evidence for an etiologic link. A probable exception to this statement is African Burkitt's Iymphoma (BL). In vitro infection of Iymphoma cells with Epstein-Barr virus ( EBV) can induce the transformation of such cells into long-term tissue culture cell lines, and the majority of patients with African Burkitt's Iymphoma have detectable antibody directed against the Epstein-Barr virus early antigen. Epstein-Barr virus infection is known to affect primarily B cells, which upon transformation are destroyed by cytotoxic T cells that respond to the EBV antigen attached to B cells. The failure of such immunologie control leads to the uncontrolled proliferation of transformed B cells, resulting in Burkitt's Iymphoma. Corresponding evidence supporting a viral etiology in the American counterpart of this disease, however, is rarely present. Antibody profiles have been normal in most children in whom viral DNA and EBV nuclear antigen have been absent. Only approximately 20% of American children with Burkitt's Iymphoma have had evidence of Epstein-Barr virus infection, and have a disease form which is indistinguishable from the African variant.

Other factors are undoubtedly involved in the pathogenesis of non-Hodgkin'slymphoma. It has recently been postulated that selective immimoregulatory defects may allow stimulation of uncontrolled proliferation of selected clones of lymphocyte cells, leading to lymphomatous transformation.2 That immunodeficiency may play a key role in pathogenesis of lymphomas is suggested by the disproportionate number of neoplasms seen in patients receiving immunosuppressive therapy for other diseases, especially the occurrence of lymphomas in renal allograft recipients. Immunodeficiency states, such as WiskottAldrich syndrome, combined variable immunodeficiency disease and ataxia-telangiectasia,…

The heterogeneous group of disorders known as nonHodgkin's Iymphoma (NHL), fortunately, is uncommonly seen in children. Although its incidence is not exactly defined, it is recognized as being one of the commonest malignancies of childhood, comprising almost 10% of childhood cancers and 60% of lymphomas. The clinical manifestations of non-Hodgkin's Iymphoma contrast strongly with Hodgkin's disease which is, forthe most part, an indolent process, occurring in a somewhat older age group and characterized by contiguous spread of the tumor. The adult and childhood forms of nonHodgkin's Iymphoma also show distinct differences. Thus, the pediatrie variant is characterized by noncontiguous dissemination, a tendency for early !eukemic conversion, mediastinal involvement, and invasion of the meninges.

Non-Hodgkin's Iymphoma is seen predominantly in the five- to 15-year age range, with a median age of eight years; it is more common in males than females by a ratio of 2:1 to 3:1. It is rarely seen under the age of one year.

The recognition of clinical features of this disease that are peculiar to children has led to recent advances in the prognosis and treatment of children with NHL. Patterns of response or of tumor behavior are now being discerned and analyzed, and the knowledge applied. Twenty years ago, the survival rate for children with non-Hodgkin's Iymphoma was under 10%. More aggressive therapeutic approaches have since been used in major centers, with marked improvement in long-term, disease-free survival.

ETIOLOGY

Type C RNA viruses and certain DNA viruses, most notably the herpes viruses, have been shown to have oncogenic potential among non-human mammalian and avian species.1 Presumably, after a period of latency, endogenous viral replication may be stimulated by exogenous exposure, such as to x-rays, chemicals, or infectious agents, leading to the development of lymphoid malignancy. Evidence to support the concept of viral oncogenesis has stemmed from the isolation of viral particles or of reverse transcriptase from tumor cells, and serologìe evidence of co-existent or pre-existent viral infection.

However, the isolation of viral material from human leukemia or Iymphoma cells is not a consistent finding, nor has there been consistent serologie evidence for an etiologic link. A probable exception to this statement is African Burkitt's Iymphoma (BL). In vitro infection of Iymphoma cells with Epstein-Barr virus ( EBV) can induce the transformation of such cells into long-term tissue culture cell lines, and the majority of patients with African Burkitt's Iymphoma have detectable antibody directed against the Epstein-Barr virus early antigen. Epstein-Barr virus infection is known to affect primarily B cells, which upon transformation are destroyed by cytotoxic T cells that respond to the EBV antigen attached to B cells. The failure of such immunologie control leads to the uncontrolled proliferation of transformed B cells, resulting in Burkitt's Iymphoma. Corresponding evidence supporting a viral etiology in the American counterpart of this disease, however, is rarely present. Antibody profiles have been normal in most children in whom viral DNA and EBV nuclear antigen have been absent. Only approximately 20% of American children with Burkitt's Iymphoma have had evidence of Epstein-Barr virus infection, and have a disease form which is indistinguishable from the African variant.

Other factors are undoubtedly involved in the pathogenesis of non-Hodgkin'slymphoma. It has recently been postulated that selective immimoregulatory defects may allow stimulation of uncontrolled proliferation of selected clones of lymphocyte cells, leading to lymphomatous transformation.2 That immunodeficiency may play a key role in pathogenesis of lymphomas is suggested by the disproportionate number of neoplasms seen in patients receiving immunosuppressive therapy for other diseases, especially the occurrence of lymphomas in renal allograft recipients. Immunodeficiency states, such as WiskottAldrich syndrome, combined variable immunodeficiency disease and ataxia-telangiectasia, have all been reported to be associated with a higher than expected frequency of malignancy, most often of the lymphoreticular type.

The increased risk of malignancy in immunodeficiency diseases may stem from detectable chromosomal aberrations. Chromosomal gaps, breaks, fragments, and dicentric and marker chromosomes are characteristic of ataxia-telangiectasia, Bloom's syndrome, and Fanconi's anemia, in which a predisposition to malignancy is evident. Additional support for the hypothesis of chromosomal defects as a factor in later malignant transformation is the consistent, though not constant, involvement of chromosome 14 in neoplasms of predominantly B cell origin. A notable example of this is the presence of abnormalities of chromosome 14 in African Burkitt's lymphoma.

Table

TABLE 1CONDITIONS PREDISPOSING TO NON-HODGKIN'S LYMPHOMA

TABLE 1

CONDITIONS PREDISPOSING TO NON-HODGKIN'S LYMPHOMA

In summary, we can conclude that genetic and environmental factors, and possibly infectious agents as well, all play a role in the genesis of this neoplasm whose origin is primarily in lymphocytes (Table 1).

CLINICAL PRESENTATION

The majority of children with non-Hodgkin's lymphoma present with generalized disease, whose origin is primarily in the lymphoid system. Neoplastic infiltration may be found in the thymus, Waldeyer's ring, Peyer's patches, appendix, or abdominal lymph nodes. About one-third of the patients present with enlargement of supradiaphragmatic lymph nodes, especially the cervical, supraclavicular, and axillary nodes (Figure 1). Such a presentation is often accompanied by anterior mediastinal involvement, which may be accompanied by pleural effusion (Figure 2). Abdominal tumors are seen in 30% to 40% of individuals, often located in the ileocecal region, mesentery, ovaries, or retroperitoneum (Figure 3). An interesting association is the increased incidence intraabd omina 1 involvement in boys, with a male:female ratio of 4:1. In fact, non-Hodgkin'slymphoma of the intestine occurs almost exclusively in males, while intra-abdominal disease in females is usually found in the ovaries or pelvic organs. The origin of lymphomas in extra lymphatic sites, such as in skin, bone, breast, orbit, parotids, testicles, rib, muscle or subcutaneous tissues is infrequent. Although approximately 30% of children with non-Hodgkin's lymphoma may develop involvement of the central nervous system (Figure 4) during the course of their illness, a minority (less than 10%) of these patients have CNS involvement at the time of diagnosis.

Children with non-Hodgkin's lymphoma may be brought to the attention of the pediatrician for numerous reasons. Complaints may be nonspecific, such as fever, anorexia, nausea, or weight loss. Symptoms may also vary with site of primary disease. Intra-abdominal disease may be signaled by vomiting, abdominal pain or distension (which may mimic the presence of appendicitis or intestinal obstruction). In children over six years of age, the finding of intussusception as the cause of abdominal pain should initiate a search for lymphoma. Dyspnea, coughing, wheezing, substernal or pleuritic pain, or respiratory distress may indicate the presence of an anterior mediastinal mass, which occasionally may produce a superior vena cava syndrome. Lymphomas originating in the skeleton may be accompanied by bony pain, or localized swelling and tenderness. Likewise, nasal congestion, rhinorrhea, epistaxis, headache and proptosis may accompany tumors originating in the nasopharynx, paranasal sinuses, or orbit.

Figure 1. Cervical lymphadenopathy in a child with nonBurkitt's lymphoma.

Figure 1. Cervical lymphadenopathy in a child with nonBurkitt's lymphoma.

Burkitt's Lymphoma

A discussion of Burkitt's lymphoma must entail description of both the African and the American variants, which appear histologically identical but differ in clinical features. This neoplasm comprises over 50% of childhood malignancies in areas of Africa where it is endemic. It shows a tendency to involve the jaw or orbit, as well as extranodal visceral sites, such as kidneys, bowel, mesentery and ovaries. Central nervous system involvement is a frequent complication whereas leukemic transformation is uncommonly seen. The median age of incidence is seven years (range two to 40 years).

Figure 2. Chest x-ray showing mediastinal adenopathy and pleural effusion in a child with non-Hodgkin's lymphoma.

Figure 2. Chest x-ray showing mediastinal adenopathy and pleural effusion in a child with non-Hodgkin's lymphoma.

In the United States, Burkitt's lymphoma is more likely to effect an older age group (median age 12 years). The American variant most often arises primarily in the abdomen; it involves extralymphoid sites such as the bone marrow to a greater extent, and it involves the central nervous system less commonly than the African variant.

Despite these clinical differences between the African and American variants of Burkitt's lymphoma, the response of both groups of patients to chemotherapy, particularly to the alkylating agents, are similar(Table 2). There should be rapid staging and initiation of therapy, following stabilization of the patient hematologically and metabolically. With chemotherapy consisting of high dose intermittent cyclophosphamide, as well as other agents, such as vincristine, methotrexate, adriamycin, prednisone, and cytosine arabi noside, superior results are being achieved. The patient who has attained complete remission can expect a prolonged survival rate of 50%. 3 Metabolic complications may arise due to rapid cell turnover and treatment-induced tumor cell lysis, the socalled acute tumor lysis syndrome. This complication may account for a significant mortality and morbidity.4 Therefore, it is essential to perform biochemical determinations, such as serum electrolytes, calcium, phosphorus, and uric acid, and frequent urinalyses. Chest radiographs, biochemical tests of liver and kidney function, and the erythrocyte sedimentation rate should also be performed. More detailed and sophisticated clinical investigations are rarely indicated because of rapid progression of the disease.

Figure 3. Gastrointestinal contrast (barium) x-ray showing involvement by tumor in a child with non-Hodgkin's lymphoma. Narrowing of the lumen of the bowel and multiple indentations are due to tymph node enlargement.

Figure 3. Gastrointestinal contrast (barium) x-ray showing involvement by tumor in a child with non-Hodgkin's lymphoma. Narrowing of the lumen of the bowel and multiple indentations are due to tymph node enlargement.

Figure 4. Lumbar myelogram showing spinal cord compression secondary to lymphoma. Flow of contrast media is completely obstructed (left).

Figure 4. Lumbar myelogram showing spinal cord compression secondary to lymphoma. Flow of contrast media is completely obstructed (left).

Table

TABLE 2COMPARATIVE CLINICAL AND LABORATORY FEATURES OF AMERICAN AND AFRICAN BURKITT'S LYMPHOMA

TABLE 2

COMPARATIVE CLINICAL AND LABORATORY FEATURES OF AMERICAN AND AFRICAN BURKITT'S LYMPHOMA

DIAGNOSIS

The advanced stage of disease at diagnosis in most children with non-Hodgkin's lymphoma demands efficient and selective diagnostic evaluation. Accurate staging or determination of extent of disease is desirable, when possible, as it allows definitive comparison and analysis of therapeutic results. However, rapid diagnosis and institution of therapy must be given the greatest priority.

The Ann Arbor staging system has been used in several series of cases of non-Hodgkin's lymphoma, and correlation of staging with survival has been demonstrated. However, the use of the Ann Arbor staging system is based upon studies of Hodgkin's disease, with its unicentric origin and contiguous spread. The noncontiguity of spread of disease, the frequency of extra nodal (bone marrow and CNS) involvement, and the rarity of a true stage III in children with non-Hodgkin's lymphoma indicates the need for a modified staging system tailored to the needs of the child with nonHodgkin's lymphoma. Efforts have been made by various centers to devise a staging scheme which could be universally accepted (Table 3).

Generous tissue blocks should be secured from the involved lymph node or tumor mass. The diagnostic evaluation should include: a complete blood count, examination of the peripheral blood smear (Figure 5), serum electrolytes, biochemical tests of liver and kidney function, and urinalysis. Bone marrow examination should also be performed, ideally with smears obtained from multiple sites. Bone marrow biopsy, which is of proven value for adult patients, has been recommended by some investigators for the pediatrie lymphoma patient. Others have claimed that the bone marrow biopsy rarely shows bone marrow involvement by lymphoma in children whose bone marrow smears obtained by aspiration are normal and, therefore, bone marrow biopsy should not be routinely performed.5 Examination of the cerebrospinal fluid is essential to rule out CNS involvement at the time of diagnosis.

Further studies may be performed, as dictated by the clinical manifestations of each individual patient. Intravenous pyelography and ultrasonography is helpful to reveal renal lymphomatous involvement or ureteral compression secondary to enlarged retroperitoneal lymph nodes. Computerized tomography and ultrasonography allows non-invasive abdominal examination. Radionuclide scans may be useful in the detection of pathology in liver, spleen or bones. Gallium scanning is of little usefulness, as are lymphangiograms, which have very little usefulness in children with non-Hodgkin's lymphoma. Lymphangiography may even be contraindicated in those patients with respiratory impairment because of the possibility of lipid embolization and pneumonitis.

Table

TABLE 3CRITERIA FOR STAGING SCHEMAS FOR CHILDHOOD NON-HODGKIN'S LYMPHOMA

TABLE 3

CRITERIA FOR STAGING SCHEMAS FOR CHILDHOOD NON-HODGKIN'S LYMPHOMA

Figure 5. Smear of peripheral blood showing presence of cells indistinguishable from acute lymphoid leukemia in a child with non-Hodgkin's ty/nphoma and conversion to leukemia (Wright's stain, magnification 1200X).

Figure 5. Smear of peripheral blood showing presence of cells indistinguishable from acute lymphoid leukemia in a child with non-Hodgkin's ty/nphoma and conversion to leukemia (Wright's stain, magnification 1200X).

Laparotomy is not routinely performed except in the child with an acute abdominal presentation. Although laparotomy enables the stage upgrading of a substantial number of patients, such a change in stage does not clearly alter the prognostic outlook or the therapeutic approach.

Thus, the diagnostic workup of a child with nonHodgkin's lymphoma is dictated by expethency (Table 4). Since early dissemination of disease is a major problem, staging may have to be deferred to allow the earliest possible institution of treatment.

HISTOPATHOLOGIC CONSIDERATIONS

Until recently, the Rappaport system of histopathologic classification of non-Hodgkin's lymphomas was most widely accepted.6 Its histologie subgroups include: nodular or diffuse undifferentiated, histiocytic, welldifferentiated and poorly differentiated lymphocytic variants, as well as the mixed histiocytic-lymphocytic type. Among children, the diffuse forms predominated, although a nodular pattern is occasionally seen in the histiocytic subtype. The mixed pattern is seen infrequently in children. Diffuse poorly differentiated, diffuse undifferentiated, and diffuse histiocytic lymphomas are the most common histologie subtypes of childhood nonHodgkin's lymphomas.

The Ra p pa port classification system relies mainly on architectural differences in the involved nodes. Thus, it has been criticized for its failure to consider functional differences among these tumors; the lack of consistent reproducibility of the histologie subgroupings using this system is also a problem. Two newer classification systems have attempted to utilize cytologie type, basing their grouping of disease upon functional criteria. These are the Lukes-Collins and Kiel classifications.6 Most lymphoproliferative disease in humans is of B cell derivation, with a smaller proportion of T cell types in children. It has been determined that the site of B cell transformation is in the follicular centers of lymph nodes. Derepression of B cell transformation may lead to proliferative changes in the lymphocyte, finally evolving into formation of a Iymphoma. What was previously termed histiocytic Iymphoma is now considered a Iymphoma of transformed lymphocytes. Both undifferentiated Burkitt's and non-Burkitt's types of Iymphoma are also derived from small, transformed, or non-cleaved follicular center cells.

Included in both the Rappaport and Lukes-Collins classifications is a Iymphoma that is termed undifferentiated lymphoblastic Iymphoma or convoluted T cell Iymphoma, respectively. This variant occurs in the somewhat older child, who is usually male. It most often involves the mediastinum and has a strong tendency towards leukemic conversion. T cell characteristics are present on immunologie marker studies.

In children, approximately 60% of cases of nonHodgkin's Iymphoma are of B cell derivation.7 Occasionally, patients may have no detectable cell surface markers and thus are of null cell origin. An even smaller percentage of children have lymphoblasts with both T cell and B cell markers, which are referred to as "D cells," whose origin and significance are unknown.8

The correlation of histopathologic type to survival is not clearly defined. However, in a recent study, although histology did not appear to influence survival, it was observed that the length of complete remission was greater for children with diffuse histiocytic Iymphoma compared to those with diffuse poorly differentiated lymphocytic Iymphoma.9 However, there is still a need for developing a system which will correlate prognosis with immunologie, cytochemical, and histopathologic features seen in pediatrie lymphomas (Table 5).

TREATMENT

Before 1970, the median survival of children with nonHodgkin's Iymphoma was approximately six months. Cure was achieved in only 10% to 20% of patients, most of whom had limited dissemination of the tumor. With truly localized disease, that is, stage I and II and most often as tumors in the neck and axilla, radiotherapy alone was able to bring about a 50% cure rate (Table 6). However, the use of aggressive combination chemotherapy has had a major impact on outlook and survival, such that overall survival statistics of 50% to 80% are now being reported. Systemic chemotherapy is therefore advocated for all children with non-Hodgkin's Iymphoma, even under the most favorable circumstances, with stage I tumors. Radiotherapy has been incorporated into many treatment regimens, although care should be exercised to avoid additive toxicity. The added toxicity from radiotherapy may interfere with administration of optimal drug therapy and delay the induction of the patient into remission. Evidence that radiation therapy to bulk disease prolongs survival is not conclusive.

Table

TABLE 4DIAGNOSTIC AND STAGING STUDIES FOR WORKUP OF A CHILD WITH NON-HODGKIN'S LYMPHOMA

TABLE 4

DIAGNOSTIC AND STAGING STUDIES FOR WORKUP OF A CHILD WITH NON-HODGKIN'S LYMPHOMA

About 90% of children with non-Hodgkin's Iymphoma will enter remission when treated with current chemotherapeutic regimens. Of those failing to enter complete remission, 80% die within six months of diagnosis. Recurrent disease is usually not seen locally, but is seen at sites distant from the original location. The first two years after diagnosis are the most crucial in terms of the likelihood of relapse, and it is now accepted that a period of 24 months of disease-free survival is tantamount to cure.

The best survival rates are seen in those patients with stage I and II disease, treated with involved field irradiation and combination chemotherapy. In this group, 80% to 1 00% show a two-year disease-free survival rate. Improved survival in intra-abdominal nonHodgkin's Iymphoma has resulted from aggressive attempt at surgical removal followed by intensive chemotherapy and radiation therapy.

Wollner et al at The Memorial Sloan-Kettering Cancer Center, reported on the LSA2-L2 regimen,10'11 which combines administration of multiple chemotherapeutic agents with reduced dosages of radiation therapy. These investigators found that larger doses of irradiation to intra-abdominal sites necessitated interruption of chemotherapy and did not improve survival. This was the first report indicating the importance of systemic chemotherapy. The practice of "second-look" surgery after a period of successful chemotherapy allows nonresectable tumors to be marked with metal clips and thus delineate radiation fields. Chemotherapy is immediately reinstituted after administration of 2000 rad to the involved field given over a period of two weeks.

With such a regimen of combined chemotherapy and radiotherapy, Wollner has managed to achieve an overall disease-free actuarial survival at 70+ months of 73%.9'10 Patients with lymphomas in stages I and II had a 100% survival rate (median 75+ months). Survival rates for patients in stage III and IV were 76% and 52%, after a median observation period of 69+ months.

Table

TABLE 5MORPHOLOGY, IMMUNOTYPE AND CYTOCHEMISTRY CRITERIA USED IN CLASSIFICATION OF NON-HODGKIN'S LYMPHOMAS

TABLE 5

MORPHOLOGY, IMMUNOTYPE AND CYTOCHEMISTRY CRITERIA USED IN CLASSIFICATION OF NON-HODGKIN'S LYMPHOMAS

Table

TABLE 6SUMMARY OF MAJOR PROSPECTIVE THERAPEUTIC STUDIES

TABLE 6

SUMMARY OF MAJOR PROSPECTIVE THERAPEUTIC STUDIES

Despite its encouraging results, this regimen is quite toxic, and efforts have been made to modify the treatment plan in an effort to reduce side effects while achieving the same degree of therapeutic success. Miser et al achieved an 80% overall survival rate (64%) for combined stages III and IV, utilizing a less toxic LSAi-Li regimen.12 A recent British study of 29 children with non-Hodgkin's lymp ho ma, 76% of whom had advance disease, revealed a 65% disease-free survival for a minimum of 42 months followup. This group used a regimen consisting of treatment with cyclophosphamtde, adriamycin, prednisone, and vincristine, together with CNS prophylaxis.13 Another multiagent regimen consisted of a combination of cyclophosphamide, vincristine, and prednisone, together with intravenous and intrathecal methotrexate (COMP), plus radiation therapy. In this Children's Cancer Study Group protocol, 68% of nonHodgkin's lymphoma patients, most of whom had widespread disease at presentation, were disease-free at 24 months from diagnosis.14 Fourteen out of 22 patients placed on a maintenance chemotherapy regimen remained in complete continuous remission for a median follow-up period fo 26 months. None of the patients with "favorable" presentations, that is, those with localized, extranodal, or gastrointestinal primaries, has relapsed.

A similar study from the Roswell Park Memorial Institute has also yielded a two-year 63% disease-free survival rate, although patients with stage IV disease did extremely poorly.15 Murphy et al at St. Jude Children's Hospital published results of a study, which also utilized a regimen consisting of vincristine, high dose cyclophosphamide, prednisone and involved field radiotherapy.16 This study included randomization for the addition of adriamycin to the standard regimen in those children with advanced stages of disease. The maintenance chemotherapy regimen consisted of 6mercaptopurine and methotrexate. The overall survival achieved was approximately 50%, and involved field irradiation added to combination chemotherapy produced no apparent benefit. However, there was a clear difference in survival rate between children with localized disease, as compared with those with disseminated disease.

Mediastinal lymphomas at one time were associated with the poorest prognosis, but now have a greater than 80% two-year survival rate, since the introduction of the LSAj-L2 and other protocols.17 Radiation therapy has no proven role in the treatment of bulk disease, including mediastinal lymphoma. Thus, radiotherapy should probably be used only for treatment of children who show incomplete regression of their lymphoma, for children with acute emergencies that are producing "mass" effects, such as superior vena cavai compression, spinal cord compression, or who show progression of disease when chemotherapy is employed alone.

A preliminary report from the Children's Cancer Study Group has produced an exciting indication that chemotherapy can be curative in about 70% of children with non-localized non-Hodgkin's lymphoma, and that selecting the regimen for optimal chemotherapy is dictated by histological type.18 Children with nonlocalized lymphoblastic non-Hodgkin's lymphoma achieved a 73% three-year disease-free survival rate with the LSA-L2 regimen, as compared with 31% three-year disease-free survival with the COMP chemotherapy regimen. On the other hand, children with nonlymphoblastic non-Hodgkin's lymphoma had a 69% twoyear survival rate with COM P compared with a 26% twoyear survival following LSA-L:.18

Despite these strides forward, major problems still exist. With lengthening of survival, as in the acute leukemias, there are increasing numbers of cases of CNS involvement in patients with non-Hodgkin's lymphoma. Before the advent of chemotherapy, the incidence of CNS relapse was less than 5%. Almost a third of children with non-localized non-Hodgkin's lymphoma now develop CNS involvement at some stage in the course of their disease. Cranial irradiation, intrathecal chemotherapy, or both, have been used to prevent CNS spread, but the efficacy of these regimens has not been proven. CNS dissemination of lymphomas is rarely seen with stage I and H disease, but better means of CNS prophylaxis are needed for stages Hl and IV.

Some promising new therapeutic approaches are under investigation, including interferon therapy, monoclonal antibodies, and bone marrow transplantation (both allogeneic and autologous). Merigan et al reported six cases of B cell lymphoma treated with intramuscular injections of leukocyte interferon.19 One patient with diffuse histiocytic lymphoma, and three patients with nodular poorly differentiated lymphocytic lymphoma had demonstrable response to interferon, with two of the responses lasting over 18 months. Also, a patient with cutaneous T cell lymphoma was treated using murine hybrid orna monoclonal anti-T cell antibody.20 Although clinical remission was not accomplished, still a striking response was effected. Exciting work in the field of monoclonal antibody research and its applicability to clinical studies is ongoing.

Equally exciting are the prospects for use of bone marrow transplantation for treatment of this group of diseases. So far, bone marrow transplantation has been reserved for patients with intractable or relapsed nonHodgkin's lymphoma.21'22 Identical twins with refractory non-Hodgkin's lymphoma received syngeneic bone marrow transplants in two centers. One patient with refractory stage IV nodular poorly differentiated lymphoma achieved initial complete remission after transplantation, but later succumbed to recurrent disease 62 weeks post-transplant. However, the Seattle group performed twin-to-twin transplants in eight patients with disseminated non-Hodgkin's lymphoma who had failed when treated with conventional chemotherapy. Four out of eight patients remain in continuous complete remission for periods of 12 to 126 months without further therapy.

However, at the present time such approaches must be reserved for those patients who have failed to enter remission following treatment with conventional therapy. Thus, bone marrow transplantation can not be generally advocated for treatment of non-Hodgkin's lymphomas until the favorable results are more proven, and the procedure entails fewer risks of life-threatening complications. The current combined modality regimens can clearly be improved. Therapy must be planned to take into account the variable features of this disorder, its heterogeneous histopathology, immunology, and clinical behavior. Questions concerning pathogenesis, staging, prevention of CNS and bone marrow involvement, the interrelationship of childhood non-Hodgkin's lymphoma and acute lymphoid leukemia, and the usefulness of radiation therapy remain to be resolved. It is expected that a better understanding of these issues will have a significant impact on the results of treating nonHodgkin's lymphoma.

COMPLICATIONS

The toxicity of intensive combination chemotherapy regimens can be considerable, and design of less aggressive but equally effective therapeutic regimens are a desirable goal. As the number of those surviving the disease has increased, it has become apparent that complications of both the disease and its treatment are manifold. Such complications may be seen acutely, either at presentation or with institution of systemic therapy; other complications are only now becoming obvious with accrual of statistics on long-term survival.

Acute Toxicity

Metabolic complications account for up to 15% of deaths in children with non-Hodgkin's lymphoma. The most common metabolic derangement is renal impairment, which in most instances results from urate nephropathy but may occasionally be due to acute renal failure. Less commonly, the kidneys themselves may be infiltrated or invaded by tumor, although such a situation does not necessarily result in acute renal failure. Mechanical urinary tract obstruction may also occur, usually from uric acid calculi, or rarely from accumulation of urate precursors, xanthine or hypoxanthine. Orotic acid or orotidme excretion may also be increased in those patients who have been treated with allopurinol to prevent urate crystal precipitation.4

Hyperkalemia may occur usually within 24 hours after institution of chemotherapy. Oliguric acute renal failure, sudden lysis of tu mor ce Us, or the combination of the two, is responsible for this problem, which may be exacerbated by metabolic acidosis. Metabolic acidosis may be the cumulative result of septic shock, hypoxia, renal failure, and production of lactic acid by tumor cells.

Hyperphosphatemia alone is fairly well-tolerated, but when the phosphate ion interacts with calcium to yield hypocalcemia, through calcium phosphate deposition, profound disturbances such as tetany, seizures, and ventricular arrhythmias may result in permanent organ damage or even death. The pathogenesis of hypercalcemia is less well understood. Since it may be observed with or without bone involvement, it obviously cannot always be attributed to bone résorption. Possible reasons for hypercalcemia are the presence of increased levels of circulating immunoreactive parai hormone due to secondary hyperparathyroidism, presence of an ectopie para t hormone-like substance, or the production by tumor cells of prostaglandin E2 or an osteoclastactivating factor.

Hyponatremia occasionally develops, probably caused by forced diuresis and fluid overload, which may lead to hypertensive changes as well. The physician administering the initial chemotherapy should be aware of these and other metabolic hazards. Examples are the induction of inappropriate ADH secretion (SIADH) with vincristine usage and retention of water secondary to the accumulation of cyclophosphamide metabolites. Hyponatremia may also be exacerbated by gastrointestinal losses of sodium from emesis or from fluid retention caused by renal failure.

Hypoglycemia and hyper lipide mia may also be present. The pathogenesis of these conditions is not well understood, although most likely they are due to tumorrelated gluconeogenesis.

The occurrence of any one of these metabolic complications requires prompt recognition and appropriate therapy. Use of allopurinol, forced diuresis and alkalinization will usually prevent or minimize the development of urate nephropathy. Close monitoring of blood pressure and fluid status will also lessen the risk of hypertensive changes. The patient with hyperphosphatemia may require simultaneous administration of calcium and reduction of serum phosphate. The patient with the opposite problem of hypercalcemia may show little or no response to forced diuresis; instead he may respond only to the use of agents such as corticosteroids, mithramycin, or thyrocalcìtonin, or to the performance of hemodialysis.

Prevention of renal failure and metabolic complications is essential in managing patients with nonHodgkin's lymphoma where many of these complications may be of life-threatening severity.

Chronic Toxiclty

Impairment of normal growth results from the combination of chronic illness and intensive chemotherapy/radiotherapy regimens as are used in children with cancer. With chemotherapy, such growth retardation is temporary; growth is usually resumed after chemotherapy is discontinued. However, the effects of radiotherapy on growth are more complex, depending to a large extent upon age, dose, treatment regimen and the timing of irradiation.

In addition, irradiation has been associated with demyelination, transverse myelitis and peripheral neuropathy. Diminished growth hormone levels have also been reported in children who were given irradiation to the head and neck region. Pneumonitis, pericarditis, and pleurisy have all been noted as types of radiationinduced effects. Recently, it has been observed that individuals who have received head and neck irradiation may become hypothyroid. The incidence of this complication may be further increased if the patient has also received lymphangiography, because of the increased uptake of iodine by the thyroid gland following injection of the radi opaque substance.

Both rad io therapeutic and chemotherapeutic modalities may reduce a patient's reproductive capacity. The dose of radiotherapy which will result in sterility of the prepubescent child is not known. In older individuals, however, it is a recognized fact that aspermia and ovarian dysfunction result from radiotherapy, and that chemotherapy is also associated with hormonal dysfunction and germinal aplasia. It is not yet known to what extent these findings are reversible, but the pubertal child is probably particularly vulnerable.

Other complications related to the therapy of the child with non-Hodgkin's lymphoma are those of possible teratogenesis, h e pa to ce Uu Ia r dysfunction, radiationinduced nephritis in those receiving doses of radiotherapy to the abdomen of more than 2000 rad, and intercurrent infections associated with therapeutically-induced neutropenia and immunosuppression. Reports of the long-term effects of chemotherapy and radiotherapy on immunocompetence, at present, are conflicting.

Of concern is the possible effect of therapy on the later development of secondary malignancies. In adults, the frequency of a secondary malignancy may approach 10%. However, whether these statistics will later change or are different in the pediatrie population must be determined after further study of accumulated data on long-term survivors.

Thus, the adverse effects of both the disease and its treatment can be considerable. New complications will probably come to light as increasing numbers of children survive their primary disease. To minimize these difficulties, thought must be given to individualizing treatment in accordance with the manifold features of the disease, and to the use of measures for early recognition and prevention of adverse side effects. In this way the prospect of increased life expectancy from successful therapy will not be tempered by an impaired quality of life.

SUMMARY AND CONCLUSIONS

Non-Hodgkin's lymphoma in the child is distinguished from its adult counterpart by a number of clinical and pathologic features, including its noncontiguity, tendency toward Jeukemic conversion, and mediastinaJ and central nervous system involvements. Pathologically, the diffuse forms, poorly differentiated, undifferentiated and histiocytic, predominate, and the majority of tumors are of Bcell derivation. Peculiar to the pediatrie age group is the undifferentiated lymphoblastic lymphoma or convoluted T-cell lymphoma, which is primarily seen in the adolescent male, with involvement of the mediastinum and a strong proclivity toward leukemic conversion.

Such differences may be of paramount importance. Attempts are currently being made to determine the correlation of histopathologic type to survival. That selection of chemotherapeutic regimen can be predicated upon histologie type is implied from recent Children's Cancer Study Group data. Generally, however, it has been established that aggressive systemic chemotherapy is necessary to achieve long-term survival and cure in the pediatrie non-Hodgkin's patient, even in those with truly localized disease. With involved field irradiation and combination drug therapy, in the treatment of patients with stage I and II disease, 80% to 100% two-year diseasefree survival rates can be achieved. Variations of the LSAi-Lz regimen from the Memorial Sloan-Kettering Cancer Center have yielded overall survival figures of 70% to 80% at 70+ months, while still other approaches, eg COMP, have yielded no less than a 63% over-all twoyear disease-free survival rate. It is now accepted that a period of 24 months of disease-free survival is probably equivalent to cure.

Yet, data on short and long-term complications resulting from intensive combination chemotherapy programs is accumulating continually. Metabolic complications, some of which may be life-threatening, may occur soon after institution of therapy. Liver, kidney, lungs, etc. may all bear the effects of various drugs employed in the therapy. Although growth usually is only temporarily impaired with the use of che mo therapeutic agents, combinations of chemotherapy and irradiation may lead not only to permanent growth retardation, but also to hormonal imbalance, as a by-product of hypothalamic-pituitary axis, thyroid or sexual organ damage. The effects on the latter organ system have also been manifested by sterility and failure of sexual maturation.

The complications of therapy and of disease are numerous, and alarming, at times. However, the most frightening of these is the occurrence of secondary malignancies in those individuals who have achieved cure. The exact incidence of this side effect in the pediatrie population is not yet known, but may approximate at least 10%.

Efforts are being made constantly to survey patients with non-Hodgkin's lymphoma to assess the incidence and impact of any and all complications these patients might incur, with an eye to prevention. Individualization and modification of therapy offers a chance for minimizing these risks, while still bringing about further improvement in survival. The immunology and pathology of the tumor are being probed for clinical and therapeutic correlates. Issues such as central nervous system relapse, or defining the role of irradiation continue to present a challenge. Novel approaches to therapy, such as Interferon therapy, monoclonal antibodies, and bone marrow transplantation are exciting treatment concepts or approaches for patients who have failed conventional treatment plans. It is expected that refinement of these techniques may afford additional hope for the individual with non-Hodgkin's lymphoma.

REFERENCES

1.Gallo RC. Meysken FL: Advances in the viral etiology of leukemia and lymphoma. Semin Hemaiot 1978; 15:379098.

2. Louie S, Schwanz RS: Immunodeficiency and the pathogenes of lymphoma and leukemia. Semin H em mol 1978; 15:117-138.

3. Zicglef JL, et al Cure of Burkilt's lymphoma; ten-year follow-up of 157 Ugandan patients. Lancei 1979: 2:936-938.

4. Lynch RE. et al: Renal and metabolic complications of childhood NonHodgkin's lymphoma. Semin Oncol 1977; 4:325-334.

5. Murphy SB: Prognostic features and obstacles to cure of childhood NonHodgkin's lymphoma. Semin Oncol 1977; 4:265-271.

6. Nathwani BN, et al: Non-Hodgkin's lymphomas; a clinicopathologic study comparing two classifications. Cancer 1978; 41:303-325.

7. Bowman WR, M el vi n S. Mauer AM: Cell markers in lymphomas and kukemias. Adv Intern Med 1980; 25:391-425.

8. Barrett SC, et al: Ly mp ho blas t s with both T and B markers in childhood leukemia and lymphoma. Blood 1979; 50:71-79.

9. Sullivan MP, et al: Pediatrie oncology group trial of LSAi-L2 therapy in Non-Hodgkin's lymphoma (N HL). Proceedings of the American Society of Clinical Research 1981; 22:378.

10. Wollner N. Exelby PR, Lieberman PH: Non-Hodgkin's lymphoma in children. Cancer 1979; 44:1990-1999,

11. Wollner N, et al: Improved prognosis in children with intra-abdominal nonHodgkin's lymphoma following LSAj-L? protocol chemotherapy. Cancer 1980; 45:3034-3039.

12. Miser JS, et al: Seventy percent survival in childhood non-Hodgkin's lymphoma; repon of 21 cases treated with a less-toxic LSAi-L; regimen. Am J Pediatr Hemaiol Oncol 1980; 2:317-320.

13. Goldman A, et al: The long-term outlook for children treated for NonHodgkin's lymphomas; a report of the children's solid tumor group. Br J Cancer 1981; 44:872-878.

14. Meadows AT. el al: A ne* therapy schedule iot pediatrie Non-Hodgkin^ lymphoma toxiciiy and preliminary results. Med Pediair Oncol I980;8:I524.

15. Brecher ML. et al: Non-Hodgkin's lymphoma in children. Cancer 1978; 41:1997-2001.

16. Murphy SB. Hustu HO: A randomized trial of combined modality therapy of childhood Non-HodgkiiA lymphoma. Cancer 1980; 45:630-637.

17. Weinstein JH, et al: Improved prognosis for patients with medias t i nal lymphoblastic lymphoma. Blood 1979; 53:687-694.

18. Jenkin RD. et al: Pediatrie Non-Hodgkin's lymphomas; the Children's Cancer Study Group Experiences; an Interim Repon. Front Radial Ther Oncol 1982: 16:134-140.

19. Merigan TC, el at: Preliminary observations on ihe effect of human leukocyte interferon in Non-Hodgkin's lymphoma. jV Engl J Med 1978; 299:1449-1453.

20. Miller RA, Levy R: Response of cutaneous T-CeIl lymphoma to therapy with hybridoma monoclonal antibody. Lancei 1981; 2:226-230,

21. Goldman J M, et al: Identical twin marrow transplantation for patients with leukemia and lymphoma. Transplantation 1981; 31:140-141.

22. Appelbaum FR. et al: Treatment of Non-Hodgkin's lymphoma with marrow transplantation in identical twins. Blood 1981; 58:509-513.

TABLE 1

CONDITIONS PREDISPOSING TO NON-HODGKIN'S LYMPHOMA

TABLE 2

COMPARATIVE CLINICAL AND LABORATORY FEATURES OF AMERICAN AND AFRICAN BURKITT'S LYMPHOMA

TABLE 3

CRITERIA FOR STAGING SCHEMAS FOR CHILDHOOD NON-HODGKIN'S LYMPHOMA

TABLE 4

DIAGNOSTIC AND STAGING STUDIES FOR WORKUP OF A CHILD WITH NON-HODGKIN'S LYMPHOMA

TABLE 5

MORPHOLOGY, IMMUNOTYPE AND CYTOCHEMISTRY CRITERIA USED IN CLASSIFICATION OF NON-HODGKIN'S LYMPHOMAS

TABLE 6

SUMMARY OF MAJOR PROSPECTIVE THERAPEUTIC STUDIES

10.3928/0090-4481-19830401-05

Sign up to receive

Journal E-contents