Cancer in Children: Clinical Management, 5th Edition

Chapter 14. Non-Hodgkin lymphoma

Catherine Patte


Lymphomas are a heterogeneous group of malignant proliferations of lymphoid cells at various stages of differentiation and activation. Recent biological studies now allow a better classification.1 In children, a few subtypes of non-Hodgkin lymphoma (NHL) are seen, predominantly the lymphoblastic and Burkitt types. They are characterized by the predominance of extranodal disease and by very rapid growth and dissemination, especially in the bone marrow and central nervous system (CNS). The prognosis of childhood NHL has changed in the last 20 years as a result of improved understanding of the pathological process and better use of intensive chemotherapeutic regimens.


NHL has a peak incidence between 7 and 10 years of age, and is uncommon before 2 years of age. There is a male predominance, especially in Burkitt lymphoma, with a ratio of 3:1. An annual incidence of approximately 7 cases per million children makes it the third most common childhood malignancy.

A few familial cases of lymphoid malignancies are observed, without recognized genetic abnormalities. Patients at increased risk of developing NHL include those with congenital immunodeficiency (such as ataxia–telangiectasia or Wiskott–Aldrich syndrome), those with AIDS, and those receiving immunosuppressive therapy (such as after organ transplantation). There is an X-linked lymphoproliferative syndrome characterized by a particular sensitivity to Epstein–Barr virus (EBV) and the occurrence of malignant lymphomas, especially of the Burkitt type, and fatal infectious mononucleosis. Specific geographical areas are also associated with particular types of lymphoma, such as the ‘endemic’ (African) Burkitt lymphoma.

Aetiology has been particularly linked with EBVafter the observation that EBV particles were present in the nucleus of malignant cells in endemic Burkitt lymphoma. Although the virus is known to have transforming and immortalizing properties on B cells in vitro and is suspected to have a role in the oncogenic process, evidence of other events is necessary to establish malignancy. Furthermore, primary evidence of viral participation in oncogenesis is rare (EBV is found in only 20 per cent of non-endemic Burkitt lymphoma) and to date no viral particles have been found in childhood T cell lymphoma.



The classification of NHL has changed many times over the years and has become clearer with understanding of the differentiation pathways of normal lymphocytes and the explosion of new diagnostic tools (immunophenotyping, cytogenetics, molecular biology, and now gene sequencing). Only four categories of lymphoma are usually encountered in children: Burkitt, lymphoblastic, large B-cell, and anaplastic large cell.1 Their characteristics are given in Table 14.1. The indicated frequency is an average in Western Europe, which depends on geography. Burkitt lymphoma represents the majority of lymphoma in equatorial Africa and is less frequent in Northern Europe and in Japan. The frequency of large cell lymphomas increases among adolescents.

Microarray technologies, by studying the expression of many genes at once, are very promising, but the implications for diagnosis and prognosis, and their further usefulness in clinical practice, especially in childhood NHL, still have to be demonstrated.

Cytogenetics and molecular biology

Cytogenetic and molecular biological studies have contributed to a better understanding of the fundamental processes in NHL, but have not yet contributed to therapy. In Burkitt lymphoma, tumour cells are characterized by a translocation involving the long arm of chromosome 8, region q24. The majority of cases exhibit the translocation t(8;14)(q24;q32). Variant translocations include t(2;8)(p12;q24) and t(8;22)(q24;q11). In translocation t(8;14), the oncogene c-myc moves from its normal position on chromosome 8 and is rearranged with the gene for heavy-chain immunoglobulin, while in the variant translocations c-myc remains on chromosome 8, and the κ and λ l immunoglobulin light-chain genes translocate from their normal positions on chromosomes 2 and 22, respectively, to a region distal to the c-myc oncogene. This results in a transcriptional deregulation of the c-myc gene and its overexpression, which is thought to play a crucial role in the genesis and/or maintenance of this malignancy.

In T-cell tumours, chromosomal abnormalities are rarer and more heterogeneous. They can involve chromosome 14, at the locus of the gene for theα chain of the T-cell receptor, or chromosome 7, at the locus of the gene for the β chain of the T-cell receptor.

Clinical features

The clinical presentation of NHL in childhood is varied and depends on the primary site of the disease, the histological subtype, and the extent of the disease.

The abdomen is the most frequent primary site (30–45 per cent of all cases). Intussusception, leading to the discovery of a small resectable abdominal tumour, is a rare presentation; a large and rapidly growing abdominal mass, often associated with ascites, is the usual clinical presentation. Ultrasound scan defines the gut and/or mesenteric mass and any intra-abdominal spread. Laparotomy should be avoided, unless there is an abdominal emergency, and the diagnosis should be made by cytological examination of ascites or percutaneous needle biopsy of the tumour. Cyto/histology shows Burkitt lymphoma in the majority of cases; others are large B-cell.

Mediastinal tumours (25–35 per cent) are typically T-cell lymphoblastic lymphomas (thymic origin), but in rare cases they can be large B-cell (thymic origin) or Burkitt lymphoma (nonthymic origin). Mediastinal compression and/or cervical or axillary lymphadenopathy is an indication for a chest radiograph which will show a mediastinal mass, often associated with pleural or pericardial effusions. Patients are at particular risk of developing respiratory distress which is worsened or provoked by general anaesthesia, which therefore must be avoided. Diagnosis should be made using cytological examination of effusions or bone marrow smears.

Table 14.1. Different types of childhood lymphomas and their biological Specifics


Burkitt (50 per cent)

Lymphoblastic (30 per cent)

Large B-cell (10 per cent)

Anaplastic large cell (10 per cent)

Preferential tumour site

Abdomen Head and neck

Mediastinum (T) Bone, (sub)cutaneous (B)

Abdomen, thymus, bone

Node, skin

Corresponding stage of F cell development

Germinal center


Germinal center

Mature T-cell


Cell size






Narrow, basophilic with vacuoles

Narrow, pale


Abundant and clearerythrophagocytosis



Sometimes convoluted

Sometimes cleaved, vesicular

Irregular, clear


Several nucleoli

Poorly discernible

Distinct, often adherent to nuclear membrane



Coarse and irregular

Finely stippled



FAB equivalent


L1, L2








B lineage T lineage




CD19+ CD7+

S Ig +/

ALK+ (T or ‘null’ markers)

Ki 67+, >95 per cent

CD 79a+ CD2+

bcl6 +/-


S Ig CD3c+

Ki 67, 60 per cent-90 per cent





t (8;14) (q24;q32) or variant

No specific abnormalities


t(2;5)(p23;q35) or variant

t(2;8) (p11;q24)

Sometimes involvement of T-cell

t(8;14) (q24;q32)

NPM/ALK fusion protein

t(8;22) (q24;q11)

antigen receptor genes (TCR) on

der (3)(q27) (bcl6)

ALK is a tyrosine kinase receptor


chromosome 7(q34) or 14(q11)


ALK located on 2p23


Transcriptional deregulation of c-myc


Transcriptional deregulation of bcl6



If a tumour biopsy is needed, this should be done by percutaneous needle biopsy or mediastinoscopy.

The third most frequent site is the head and neck (10–20 per cent), including Waldeyer's ring and the facial bones, followed by the superficial lymph nodes (5–10 per cent). The remaining 5–10 per cent include tumours that arise from less common sites such as bone, skin, thyroid, orbit, eyelid, kidney, and epidural space. Diagnosis is confirmed by tissue biopsy with immunophenotyping since no particular type of lymphoma is associated with these sites, except Burkitt with facial bones and orbit, especially in endemic Burkitt NHL. Bone lymphoma may be localized, but can be also generalized and is often associated with hypercalcaemia. Kidney lymphoma can be confused with nephroblastoma, but in lymphoma the tumour is frequently bilateral, the infiltration multinodular or diffuse, or renal failure may be present. (Sub)cutaneous lymphoma particularly occurs in young children aged <2 years and is of precursor B-cell phenotype.

Anaplastic large-cell lymphomas have clinical particularities. Nodal involvement, sometimes painful, is characteristic of this disease; other characteristics are frequent skin involvement with inflammatory symptoms in the involved nodes, distant macular lesions, or general skin changes resembling ichthyosis, and frequent generalized symptoms with widely fluctuating fever. A few cases show a ‘wax and wane’ pattern with previous episode(s) of spontaneous regression.

Initial workup and staging

Diagnosis can be made using biopsy material, including tumour-touch preparations, or cytological examination of effusion fluid or bone marrow smears. Thus surgical procedures can be avoided in diffuse disease. Immunological studies are recommended, preferably obtained on suspended cells from fresh tumours or on frozen sections, but immunohistochemistry on paraffin sections is also of value. Cytogenetic studies are also important. Once the diagnosis of NHL has been made, a speedy assessment of diagnosis, staging, and general evaluation must be performed in order to commence appropriate treatment as soon as possible.

Because of the predominance of extranodal primaries and the unpredictable pattern of spread, the Ann Arbor classification is not suitable for childhood NHL. Several classifications have been proposed, but the system most commonly used is the St Jude staging system (also known as the Murphy system) (Table 14.2). The traditional boundary between leukaemia and lymphoma has been arbitrarily defined as more or less than 25 per cent blast cells in the bone marrow, but this does not correspond to either clinical or biological differences. CNS involvement is defined by the presence of unequivocal malignant cells in a cytocentrifuged specimen of spinal fluid and/or obvious neurological deficits such as cranial nerve palsies. Although not included in the staging system and not a specific marker, lactate dehydrogenase (LDH) level in the serum is a very good indicator of tumour burden and often has a prognostic significance.

Workup procedures are outlined in Table 14.3. Abdominal ultrasonography is adequate to define abdominal disease; CT scanning is not necessary in the primary investigation of extended abdominal and mediastinal disease. Experience with positron emission tomography (PET) scanning in childhood NHL is still at an early stage. It is hoped that this investigation will help to predict active tumour in a residual mass, but its role still needs to be evaluated

Patients often have other problems, such as malnutrition, infection, postsurgical complications, and respiratory and metabolic abnormalities; these may be life threatening or compromise the onset of therapy. For example, hypercalcaemia requires treatment as a separate entity. If there is renal failure, its mechanism must be established. It may be uric acid nephropathy, tumour infiltration, or urinary obstruction. In order to start appropriate treatment it is necessary to start hyperhydration promptly. Sometimes dialysis or transcutaneous pyelostomy must be performed to obtain sufficient urine outflow.

Table 14.2. St Jude's staging for childhood non-Hodgkin lymphoma

Stage I

·  A single tumour (extranodal) or single anatomical area (nodal) with the exclusion of the mediastinum or abdomen

Stage II

·  A single tumour (extranodal) with regional node involvement

·  Two or more nodal areas on the same side of the diaphragm

·  Two single (extranodal) tumours with or without regional node involvement on the same side of the diaphragm

·  A primary gastrointestinal tract tumour, usually in the ileocoecal area, with or without involvement of associated mesenteric nodes only, grossly completely resected

Stage III

·  Two single tumours (extranodal) on opposite sides of the diaphragm

·  Two or more nodal areas above and below the diaphragm

·  All the primary intrathoracic tumours (mediastinal, pleural, thymic)

·  All extensive primary intra-abdominal disease, unresectable.

·  All paraspinal or epidural tumours, regardless of other tumour site(s)

Stage IV

·  Any of the above with initial CNS and/or bone marrow involvement

The tumour lysis syndrome may be present at diagnosis or develop during treatment. Preventive measures, a diuretic and a ‘uricolytic’ drug (allopurinol or urate oxidase), must always be instituted. Urate oxidase is preferable in cases of advanced-stage disease. It transforms uric acid into allantoin, which is highly soluble in urine. It has been shown to be more efficient in promptly reducing serum uric acid level, preventing uric acid nephropathy, and preserving renal function, allowing better excretion of other cell metabolites such as potassium and phosphorus.2,3,4 The recombinant form of urate oxidase is now available. Strict clinical and metabolic monitoring of patients is essential during the lysis phase.

Table 14.3. Initial workup


·  Physical examination

·  Chest and nasopharyngeal radiography

·  Abdominal ultrasonography

·  Two BM aspirations (+ BM biopsies in large-cell NHL)

·  CSF examination

·  Complete blood count

·  LDH, serum electrolytes, BUN, creatinine, uric acid levels

Optional depending on clinical circumstances

·  Bone scan and skeletal survey

·  Local CT scan (head and neck tumours)

·  MRI (CNS disease)

·  Abdominal CT scan (stage I or abdominal stage II)

·  Thoracic CT scan

BM, bone marrow; NHL, non-Hodgkin lymphoma; CSF, cerebrospinal fluid; LDH, lactate dehydrogenase; BUN, blood urea nitrogen.

Both the French and the German protocols start with a ‘prephase’ with low-dose therapy. This generally induces good tumour regression and allows initial problems to be managed without incurring the haematological and mucous complications of the more intensive regimen which commences a week later.


Before 1970 only a few localized tumours could be cured by surgery and radiotherapy. Since then, considerable therapeutic improvements have been achieved.

Indications for surgery are very rare. A small localized abdominal tumour may be completely excised in the context of intestinal intussusception. Primary surgical excision or debulking of abdominal, head and neck, or mediastinal tumours should not be attempted. Extensive surgery is uselessly mutilating and is often followed by tumoral regrowth, as well as delaying and possibly complicating chemotherapy. Rarely, abdominal surgery is needed during treatment for complications such as intestinal perforation. At the time of remission evaluation, any residual mass should be removed or widely biopsied for pathological examination to determine whether a complete (no viable cells) or partial remission has been obtained.

Radiotherapy can be efficacious in lymphoma, but this local therapy gives no advantage over chemotherapy alone and adds both immediate and long-term toxicity. Indications for radiotherapy have become very rare. It may be used in some emergency situations, although chemotherapy is often as effective, in a few cases of localized residual tumours, or for CNS-directed therapy (cranial irradiation).

Chemotherapy is the treatment of choice in childhood NHL, which should be considered to be a systemic disease even in the presence of apparently local disease. It must be tailored to the histological subtype and the extent of the disease. CNS prophylaxis is essential in Burkitt and lymphoblastic NHL. It can be given by high dose (HD) methotrexate (MTX) and intrathecal (i.t.) injections of MTX hydrocortisone and cytosine arabinoside (Ara-C).

B-cell lymphomas

Most studies relate to Burkitt lymphoma, but in some recent European studies large B-cell lymphomas were included (where they represent 10–15 per cent of cases). Fractionated cyclophosphamide (CPM), intermediate(ID) MTX or HD-MTX, and Ara-C are the most important drugs in the treatment of Burkitt NHL. Vincristine, adriamycin, etoposide (VP16), ifosfamide, and prednisone are also effective. Drugs are administered in various combinations, but usually as short pulsed courses.

Since 1981, the largest studies have taken place in France (LMB protocols) and Germany (BFM protocols), and as a result of these well-organized multicentre studies there has been considerable improvement in the treatment of B-cell lymphomas.

LMB protocols

The general scheme of the LMB protocols is shown in Figure 14.1. A cytoreductive phase (COP course) with low doses of vincristine, cyclophosphamide, and prednisone is given 1 week before the intensive induction (two consecutive courses of COPAD M) which is based on HD-MTX and fractionated cyclophosphamide. This is followed by two consecutive consolidation courses based on Ara-C in continuous infusion. CNS prophylaxis is given by HD-MTX and intrathecal injections. The French Paediatric Oncology Society (SFOP) conducted several consecutive multicentre studies (LMB81, LMB84, LMB86, and LMB89. In the first three studies, designed for advanced stage Burkitt disease, event-free survival (EFS) increased up to 75–90 per cent for CNS-patients, whereas duration of treatment was progressively reduced from 12 to 4 months (randomized LMB84 trial)5 and toxicity, especially the toxic death rate, decreased as the experience of the investigators increased. CNS prophylaxis with HD-MTX (3 g/m2 infused for 4 h) and i.t. MTX was effective (CNS relapse rate <2 per cent). It became clear that partial remission (with documented viable cells in the residual mass) at the end of induction could be cured by treatment intensification with autologous bone marrow transplantation, and that the absence of tumour reduction after COP was indicative of a poor prognosis with an EFS of 29 per cent (LMB84). For patients with initial CNS involvement whose EFS was only 19 per cent in the LMB 81 protocol, the pilot LMB 86 study using a higher dose of MTX (8 g/m2 in 4 h), triple intrathecal injections, and consolidation with continuous infusion and high-dose Ara-C and VP16 (CYVE courses) succeeded in increasing EFS to 75 per cent.

Fig. 14.1 Different courses of LMB protocols: VCR, vincristine; ADR, adriamycin; CPM. cyclophosphamide; Pred, prednisone; Ara-C, cytosine arabinoside; MTX, methotrexate; HC, hydrocortisone; CI, continuous infusion.

All mature B-cell diseases [Burkitt and large B-cell NHL and acute lymphoblastic leukaemia (ALL)], whatever the stage, were included in the LMB89 study (1989–1996).6 Patients were stratified into three therapeutic groups: group A (resected stage I and abdominal stage II) received two courses of COPAD (vincristine, cyclophosphamide, adriamycin and prednisone). Group B (patients not eligible for groups A or C) received a 3.5-month treatment identical to the short arm of the LMB84 protocol. Group C (patients with CNS involvement and ALL with 70 per cent of blasts in bone marrow) received a more intensive treatment similar to that of the LMB86 protocol. Treatment was further intensified for patients who did not respond to COP in group B and any patient with residual viable cells after the consolidation phase. A total of 561 patients were enrolled. Five-year survival was 92.5 per cent [95 per cent confidence interval (CI), 90–94] and EFS was 91 per cent (95 per cent CI, 89–93). EFS was 98 per cent for stages I and II, 91 per cent for stage III, and 87 per cent for stage IV and B-ALL. The outcome was similar in Burkitt and large B-cell disease (EFS of 92 per cent and 89 per cent, respectively). In group B, multivariate analysis of prognostic factors showed that an LDH level more than twice the normal value N (89.5 per cent for LDH >2N versus 95 per cent for LDH 2N), no response after COP (EFS, 72 per cent) and age 15 years were associated with a lower EFS. CNS involvement was the only prognostic factor found in group C (79 per cent for CNS+ versus 90 per cent for CNS).

The next study, FAB LMB96 (May 1996–June 2001), was a randomized international trial with the participation of the SFOP, the UK Children's Cancer Study Group (UKCCSG) and the US Children's Cancer Group (CCG). It was an attempt to reduce treatment further, especially cyclophosphamide dosage, to avoid sterility in boys, to reduce its duration, and to avoid cranial irradiation in patients with initial CNS involvement. Preliminary results indicate that treatment can be reduced further in group B, but not in group C. It also indicates that delay between the two induction courses has a prognostic impact.

BFM protocols

In the four consecutive BFM studies (BFM81, BFM83, BFM86, and BFM90) treatment intensity was progressively increased, treatment duration reduced, and CNS irradiation withdrawn. In the BFM90 study (1990–1995), treatment was stratified into three risk groups (slightly different from those of the French study). It consists of two, four, or six 5-day courses including dexamethasone, MTX 5 g/m2 (or 0.5g/m2 in the lowest risk group) in 24 h infusion and intrathecal injections in each course, and alternating ifosfamide, cytarabine, and etoposide with cyclophosphamide and doxorubicin. The 6-year EFS was 89 per cent for the whole group of 413 patients, and 97 per cent, 98 per cent, 88 per cent, 73 per cent, and 74 per cent for stage I, stage II, stage III, stage IV, and ALL, respectively.7

The most remarkable result is the increase of EFS from 50 per cent to 80 per cent for patients with stage IV and L3-ALL (BFM86)8 and those with abdominal stage III and LDH > 500 (BFM90)7 when MTX was increased from 0.5 to 5 g/m2. In the following BFM 95 study, the duration of the MTX infusion was randomized (24 h versus 4 h). Final data indicate that better results were obtained in the 24-h infusion arm for patients at higher risk.

Other protocols

Other studies have reported interesting results, especially the Pediatric Oncology Group (POG) randomized multicentre trials, and the National Cancer Institute (NCI)9 and Milan Istituto Nazionaledi Tumori10single-centre series. In countries that have limited resources but where Burkitt lymphoma is frequent, therapeutic improvements have also been achieved by adapting therapy to the local environment and by better management of the initial, especially metabolic, problems.

Lymphoblastic lymphomas

Patients with lymphoblastic lymphomas must be treated with protocols similar to those used in high-risk forms of ALL. Most of these lymphomas are T-lineage, but a few are B-lineage. Although their biology is not similar, they are generally treated using the same protocols. The most frequently used protocols were the LSA2L2 or derived protocols, and the non-B BFM8,11 or derived protocols.

In the BFM protocol, induction (protocol I: prednisone, vincristine, daunorubicin, L-asparaginase followed by cyclophosphamide, Ara-C, and 6-mercaptopurine, and associated with MTX i.t.) is followed by four HD-MTX courses (5 g/m2 in 24 h infusion) and reinduction (protocol II: same drugs as protocol I, but lower total doses and dexamethasone replacing prednisone). These phases are followed by maintenance with daily 6-mercaptopurine and weekly MTX. Preventive cranial irradiation was progressively reduced and then omitted. In study 90, which enrolled 101 patients, EFS improved to 90 per cent without significant change in the chemotherapy regimen.11 This might have been due to the long experience of the investigators with the protocol. An international European randomized study based on the BFM scheme is about to start.

Other protocols

Other interesting multicentre studies have been published by the UKCCSG, the POG, and the CCG. In particular, the POG has demonstrated the effectiveness of asparaginase12 and is demonstrating that of HD-MTX.

Anaplastic large-cell lymphomas (ALCLs)

This disease is rare, and so the largest series include only about 10 patients per year. In the USA, these lymphomas have not been considered as a special entity and were included in the same protocols as the other large cell lymphomas. Interestingly, however, a POG randomized study recently showed that the addition of HD-MTX and Ara-C to the APO regimen was beneficial for large B-cell lymphoma, but not for ALCL. In Europe, ALCL (initially called ‘malignant Histiocytosis’) has been considered as a specific entity for a long time, and has been generally treated with ‘B-cell-like’ rather than ‘T-cell-like’ protocols. Overall survival rate is around 70–80 per cent.13,14,15 EFS is often lower, indicating that relapses can be salvaged. Weekly vinblastine has shown its efficacy in relapse.16 The analysis of the pooled data of the German,13 French,14 and UK series indicated that the BFM regimen was at least as effective as the others with lower total doses of drugs, especially cyclophosphamide and adriamycin. It also showed the following poor prognostic factors: mediastinal, visceral (lung, liver, spleen), and skin involvement. This analysis is the basis of the ongoing international European study ALCL 99 which uses the BFM backbone and asks two additional questions.

1. What is the best way of administrating HDMTX: 1 g/m2 in 24 h infusion with i.t. MTX or 3 g/m2 in 3 h infusion without i.t. MTX? 2. What is the role of vinblastine as maintenance therapy in higher-risk patients.

The COG is starting a study which also addresses the question of vinblastine.


What are the indications for high-dose chemotherapy and haematopoietic stem cell transplantation?

The need for high-dose chemotherapy with haematopoietic stem cell transplantation has greatly diminished in parallel with the improvement in survival using intensive conventional regimens. The indications are restricted to some initially poor responders and to relapses which respond to a second-line treatment. However, relapses that occur with the present protocols are more ‘resistant’ than previously, and finding an effective second-line chemotherapy has become a challenge.

Can granulocyte colony-stimulating factor decrease neutropenia and the related complications?

Most of the polychemotherapy regimens used to treat childhood NHL are intensive and followed by pancytopenia, especially neutropenia. Therefore there was a great deal of hope that granulocyte colony-stimulating factor (GCSF) would decrease the frequency of neutropenia and its related complications. However, three randomized studies performed in childhood NHL, one by the POG in T-lymphoblastic lymphoma,17 one by the NCI in B-lymphoblastic lymphomas,9 and one in the SFOP protocols following COPADM courses,18 did not show any clinical advantage of the use of GCSF. However, these conclusions, which are applicable in the countries where the studies were performed, might be different in other countries where socioeconomic context and bacterial environment are different.

Are there other prognostic factors within St Jude staging which could indicate patients requiring more or less treatment?

Since relatively high survival rates have now been reached, the question is raised of how to decrease further treatment intensity without jeopardizing survival rates. To do so, it is necessary to find new prognostic factors.

No-one has yet succeeded in determining the prognostic factors in lymphoblastic lymphoma. Response to corticosteroids at D8, which is such a powerful one in T-ALL, could not be demonstrated in NHL. This might be due in part to the difficulty of finding a measure equivalent to and as simple as 1000 blasts in the blood.

In B-cell disease, the absence of tumour regression 1 week after COP, which is a factor for poor prognosis in the LMB84 study, was taken into account in order to intensify chemotherapy early in subsequent studies. Outcome improved from 22 to 70 per cent in the LMB89 study,6 but B-cell disease with bad response at day 7 remained among the poor prognostic factors.

Is age a prognostic factor? Age >15 years was found to be prognostic for EFS in the LMB89 study, but not for survival, and the number of patients was small.6 Age was not prognostic in two small series of both child and adult patients with B-cell NHL at the Istituto Nazionale di Tumori, Milan. Very few data are available on patients between 15 and 20 years because they are treated in different protocols in both paediatric and adult departments, and are often not registered in studies. An effort should be made in the future to clarify this question.

With new technologies, especially DNA microarray, there is hope of identifying biological features that could emerge as independent prognostic factors. However, it should be noted that all therapeutic improvements so far have been obtained independently of understanding the biology of the lymphoma.

Treatment intensity could be further decreased for some patients. This is being done for B-cell resected stage I and abdominal stage II. In the French LMB 89 study these were treated with only two courses of COPAD, and in BFM90 resected tumours were treated with two courses of chemotherapy. Are non-resected stage I and II treated too intensively in LMB group B? However, when treatment is thought not to have immediate life-threatening consequences or long-term sequelae, the rationale behind a further decrease in therapy is debatable.

Are there also groups of patients within the advanced stages for whom treatment could be decreased, knowing that a prognostic factor identified in one study might not be relevant in another? LDH level has generally been recognized as prognostic, especially among stage III patients, but bone marrow involvement, which is prognostic in most studies, was not prognostic in the LMB studies. On the other hand, reducing further treatment might not be so easy. Attempts were made in group C of the FAB LMB96 study and in arms 3 and 4 of BFM95, but interim analyses indicated that both studies should be stopped because of inferior EFS in the experimental arm.

One of the aims of decreasing treatment is to decrease long-term sequelae such as infertility, especially in boys. Cyclophosphamide at a cumulative dose > 9g/m2 produces male sterility, and so reaching this dosage should be avoided, at least for patients who have no poor prognostic factors. In fact, it would be preferable not to exceed 5 g/m2 because of individual variation in susceptibility, and the importance of factors such as age at treatment or mode of drug administration remain undetermined.

How should post-transplant B-cell non-Hodgkin lymphoma be managed?

With the development of organ transplantation, there is an awareness of both polyclonal and monoclonal B-cell lymphomas developing in patients profoundly immunosuppressed by immunosuppressive drugs. Such patients, usually recipients of cardiac or renal transplants, may develop ‘high-grade’ tumours that are histologically indistinguishable from true malignant disease. In some patients there may be a clear viral pathogen, such as EBV, in which case a polyclonal tumour may be demonstrable.

The outcome of these tumours is generally good, and they often resolve spontaneously with reduction or cessation of ciclosporin. However, in a number of patients the disease will be more aggressive and may require therapy. Treatment with monoclonal antibodies such as anti-CD20 is indicated in B-cell lymphoproliferative disease.

What is the place of tumour-specific targeted therapy in B-cell non-Hodgkin lymphoma?

This is a challenge for the coming years. Monoclonal-antibody-targeted therapy for B-cell lymphomas in adults has produced encouraging results. The anti-CD20, rituximab, has been studied in large cohorts of patients. Its efficacy has been proved in follicular lymphoma, and in large B-cell disease in elderly patients in association with chemotherapy. It is now being studied in large B-cell lymphomas in younger adults. Except in post-transplant lymphoproliferative disease, it has not yet been evaluated in paediatric practice. Neither has it been evaluated in Burkitt lymphoma at any age. The therapy seems to have only minor side effects, but it is very expensive. Taking into account the high survival rate now obtained, its place in the treatment of Burkitt lymphoma in children will be difficult to assess. However, it is necessary to evaluate such a specific therapy, knowing that no phase II studies have been performed in this disease.


Appropriate management at diagnosis of patients with NHL is essential. Prophylaxis and treatment of tumour lysis syndrome is important. Urate oxidase is more effective than allopurinol in reducing uric acid level and preserving renal function.

NHL in children is a fast-growing tumour which disseminates widely, especially in bone marrow and the CNS. Considerable therapeutic improvement has been achieved by prospective multicentre studies. Chemotherapy is the treatment of choice and its modality differs according to the histologic subtype.

Burkitt and large B-cell NHL are treated with intensive pulsed chemotherapy where the most important drugs are cyclophosphamide, HD-MTX, and Ara-C, and the cure rate is 80–90 per cent. The place of rituximab (anti CD20) is not yet determined. Lymphoblastic NHL is treated with intensive semicontinuous and prolonged chemotherapy and the cure rate is 75–90 per cent. In both diseases, CNS prophylaxis is essential and is based on intrathecal injections (MTX Ara-C). Cranial irradiation is not necessary. The optimal treatment for ALCL is still under debate. The cure rate is 70–80 per cent.

New tools, such as microarray technology, will tell us more about the biology of these diseases, but prognostic factors identified in one study might not be relevant in another. PET scanning for evaluation of disease extension or of residual masses has yet to be evaluated.


1. Jaffe ES, Harris NL, Diebold J, Muller-Hermelink HK (1999). World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues. A progress report. Am J Clin Pathol 111 (1 Suppl 1), S8–12.

2. Patte C, Sakiroglu C, Ansoborlo S, et al. (2002). Urate-oxidase in the prevention and treatment of metabolic complications in patients with B-cell lymphoma and leukemia, treated in the Société Franc¸aise d'Oncologie Pediatrique LMB89 protocol. Ann Oncol 13, 789–95.

3. Pui CH, Mahmoud HH, Wiley JM, et al. (2001). Recombinant urate oxidase for the prophylaxis or treatment of hyperuricemia in patients with leukemia or lymphoma. J Clin Oncol19, 697–704.

4. Goldman SC, Holcenberg JS, Finklestein JZ, et al. (2001). A randomized comparison between rasburicase and allopurinol in children with lymphoma or leukemia at high risk for tumor lysis. Blood97, 2998–3003.

5. Patte C, Philip T, Rodary C, et al. (1991). High survival rate in advanced-stage B-cell lymphomas and leukemias without CNS involvement with a short intensive polychemotherapy: results from the French Pediatric Oncology Society of a randomized trial of 216 children. J Clin Oncol 9, 123–32.

6. Patte C, Auperin A, Michon J, et al. (2001). The Société Franc¸aise d'Oncologie Pédiatrique LMB89 protocol: highly effective multiagent chemotherapy tailored to the tumor burden and initial response in 561 unselected children with B-cell lymphomas and L3 leukemia. Blood 97, 3370–9.

7. Reiter A, Schrappe M, Tiemann M, et al. (1999). Improved treatment results in childhood B-cell neoplasms with tailored intensification of therapy: A report of the Berlin–Frankfurt–Munster Group Trial NHL-BFM 90. Blood 94, 3294–306.

8. Reiter A, Schrappe M, Parwaresch R, et al. (1995). Non-Hodgkin's lymphomas of childhood and adolescence: results of a treatment stratified for biologic subtypes and stage. A report of the Berlin–Frankfurt–Munster Group. J Clin Oncol 13, 359–72.

9. Magrath I, Adde M, Shad A, et al. (1996). Adults and children with small non-cleaved-cell lymphoma have a similar excellent outcome when treated with the same chemotherapy regimen. J Clin Oncol 14, 925–34.

10. Spreafico F, Massimino M, Luksch R, et al. (2002). Intensive, very short-term chemotherapy for advanced Burkitt's lymphoma in children. J Clin Oncol 20, 2783–8.

11. Reiter A, Schrappe M, Ludwig WD, et al. (2000). Intensive ALL-type therapy without local radiotherapy provides a 90 per cent event-free survival for children with T-cell lymphoblastic lymphoma: a BFM group report. Blood 95, 416–21.

12. Amylon MD, Shuster J, Pullen J, et al. (1999). Intensive high-dose asparaginase consolidation improves survival for pediatric patients with T cell acute lymphoblastic leukemia and advanced stage lymphoblastic lymphoma: a Pediatric Oncology Group study. Leukemia 13, 335–42.

13. Reiter A, Schrappe M, Tiemann M, et al. (1994). Successful treatment strategy for Ki-1 anaplastic large-cell lymphoma of childhood: a prospective analysis of 62 patients enrolled in three consecutive Berlin–Frankfurt–Munster group studies. J Clin Oncol 12 899–908.

14. Brugieres L, Deley MC, Pacquement H, A et al. (1998). CD30 (+) anaplastic large-cell lymphoma in children: analysis of 82 patients enrolled in two consecutive studies of the French Society of Pediatric Oncology. Blood 92, 3591–8.

15. Seidemann K, Tiemann M, Schrappe M, et al. (2001). Short-pulse B-non-Hodgkin lymphoma-type chemotherapy is efficacious treatment for pediatric anaplastic large cell lymphoma: a report of the Berlin–Frankfurt–Munster Group Trial NHL-BFM 90. Blood 97, 3699–706.

16. Brugieres L, Quartier P, Le Deley MC, et al. (2000). Relapses of childhood anaplastic large-cell lymphoma: treatment results in a series of 41 children. A report from the French Society of Pediatric Oncology. Ann Oncol 11, 53–8.

17. Laver J, Amylon M, Desai S, et al. (1998). Randomized trial of r-metHu granulocyte colonystimulating factor in an intensive treatment for T-cell leukemia and advanced-stage lymphoblastic lymphoma of childhood: a Pediatric Oncology Group pilot study. J Clin Oncol 16, 522–6.

18. Patte C, Laplanche A, Bertozzi AI, et al. (2002). Granulocyte colony-stimulating factor in induction treatment of children with non-Hodgkin's lymphoma: a randomized study of the French Society of Pediatric Oncology. J Clin Oncol 20, 441–8.

If you find an error or have any questions, please email us at Thank you!