|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology CNS Involvement in Children With Newly Diagnosed Non-Hodgkin’s LymphomaFrom the Departments of Hematology/Oncology, Radiation Therapy, Pathology and Laboratory Medicine, and Biostatistics, St Jude Children’s Research Hospital, and University of Tennessee at Memphis, College of Medicine, Memphis, TN; and Department of Pediatrics, Northwestern University School of Medicine, and Children’s Memorial Hospital, Chicago, IL. Address reprint requests to John T. Sandlund, MD, Department of Hematology-Oncology, St Jude Children’s Research Hospital, 332 N Lauderdale, Memphis, TN 38101; email john.sandlund{at}stjude.org
PURPOSE: To determine the frequency of CNS involvement at diagnosis of non-Hodgkin’s lymphoma (NHL), to characterize its pattern of presentation, and to determine its prognostic significance. PATIENTS AND METHODS: We reviewed the records of 445 children (1975 through 1995) diagnosed with NHL (small noncleaved cell NHL/B-cell acute lymphoblastic leukemia [SNCC NHL/B-ALL], 201 patients; lymphoblastic, 113; large cell, 119; other, 12). Tumor burden was estimated by serum lactate dehydrogenase (LDH) measurement and reclassification of disease stage irrespective of CNS involvement (modified stage). RESULTS: Thirty-six of 445 children with newly diagnosed NHL had CNS involvement (lymphoma cells in the CSF [n = 23], cranial nerve palsy [n = 9], both features [n = 4]), representing 13%, 7%, and 1% of small noncleaved cell lymphoma, lymphoblastic lymphoma, and large-cell cases, respectively. By univariate analysis, CNS disease at diagnosis did not significantly impact event-free survival (P = .095), whereas stage and LDH did; however, children with CNS disease at diagnosis were at 2.0 times greater risk of death than those without CNS disease at diagnosis. In a multivariate analysis, CNS disease was not significantly associated with either overall or event-free survival, whereas both serum LDH and stage influenced both overall and event-free survival. Among cases of SNCC NHL/B-ALL, CNS disease was significantly associated with event-free and overall survival (univariate analysis); however, in multivariate analysis, only LDH had independent prognostic significance. Elevated serum LDH or higher modified stage were associated with a trend toward poorer overall survival among children with CNS disease. CONCLUSION: A greater tumor burden at diagnosis adversely influences the treatment outcome of children with NHL and CNS disease at diagnosis, suggesting a need for ongoing improvement in both systemic and CNS-directed therapy.
THE NON-HODGKIN'S lymphomas (NHL) diagnosed in children are primarily of high-grade histology. These malignancies are more likely to involve the CNS at diagnosis in patients with advanced-stage disease (eg, bone marrow involvement), head and neck primary tumors, or the endemic subtype of Burkitt’s lymphoma.1 The St Jude staging system described by Murphy2 defines CNS involvement by the presence of malignant cells in the CSF or by cranial nerve palsy (CNP). However, some groups do not include isolated CNP as a criterion for defining CNS disease. Although CNS involvement has been thought to be associated with a poor treatment outcome, little is known about the frequency and prognostic significance of lymphoma cells in the CSF (CSF+) versus the presence of CNP (either overall or in relation to histologic type). Likewise, the impact of total-body tumor burden on treatment outcome among children with newly diagnosed NHL and CNS involvement is not well characterized.
Patients Four hundred forty-five children with newly diagnosed NHL (small noncleaved cell [SNCC] and B-cell acute lymphoblastic leukemia [ALL], n = 201; lymphoblastic lymphoma, n = 113; large-cell lymphoma, n = 119; lymphoma not otherwise specified, n = 12) were treated at St Jude Children’s Research Hospital from 1975 through 1995. Histologic classification of these NHLs was based on the National Cancer Institute Working Formulation.3 Patients were staged using the St Jude staging system as previously described.2,4 CNS involvement was identified by either malignant cells in the CSF or by CNP on physical examination. Staging work-up included chest x-ray, abdominal ultrasonography and/or computed tomography of the chest, abdomen, and pelvis; bone scan; plain films of bones that were positive on bone scan; bone marrow study (aspirate and/or biopsy); and CSF examination. A complete blood cell count and a chemistry panel including serum lactate dehydrogenase (LDH) were also determined at presentation.
Treatment
Statistical Analysis
Frequency and Sites of CNS Involvement Thirty-six (8.1%) of the 445 children with newly diagnosed NHL had CNS involvement at diagnosis. The 22 boys and 14 girls with CNS involvement ranged in age from 2 months to 18 years (median, 9 years). CNS involvement was established by the presence of lymphoma cells in the CSF in 23 patients, by CNP in nine, and by both features in four (Table 1). Cranial nerves affected were as follows: II (n = 2), III (n = 6), IV (n = 4), V (n = 1), VI (n = 3), VII (n = 9), and XII (n = 1). In six patients, multiple cranial nerves were involved. Among the nine patients with CNP, computed tomography or magnetic resonance imaging scans of brain were obtained in seven, of which two were abnormal with orbital involvement. There were no other abnormalities detected. CNS involvement was significantly associated with bone marrow and/or bone involvement but not with head and neck involvement (Table 2).
Of the 36 children with CNS disease at diagnosis, 27 (75%) had the SNCC subtype, eight (22%) had lymphoblastic lymphoma, and one (3%) had large-cell lymphoma. There was a significant association between histologic subtype and associated CNS involvement (P = .0019). The frequency of CNS involvement in large-cell NHL (1%) was significantly lower than in SNCC (13.4%, P < .0001) or lymphoblastic NHL (7%; P = .05); difference in the frequency of CNS involvement between the latter two subtypes did not attain statistical significance (P = .29).
Treatment Outcome
Pattern of Relapse Twelve of the 36 children who presented with CNS disease experienced disease recurrence (Table 4). The most frequent site of failure was the CNS (nine cases). The bone marrow was the second most common site of failure (three cases). One patient had a combined CNS and bone marrow relapse. There was no obvious relationship between the pattern of CNS disease presentation (lymphoma cells in the CSF v CNP) at diagnosis and at relapse.
Impact of Tumor Burden on Outcome Tumor burden among the 36 children was heterogeneous. To better correlate this variable with outcome, we restaged each case without reference to CNS involvement (Table 5). Those reclassified as having stage I/II disease had lower serum LDH activity than did patients who retained a stage IV classification on the basis of bone marrow involvement. Patients with intrathoracic or intra-abdominal disease (stage III) had an intermediate serum LDH activity. Notably, as reclassified, there were no deaths among the five patients with stage I/II disease, four deaths (36%) among the 11 with stage III disease, and 12 deaths (60%) among the 20 with stage IV disease, representing a trend toward a poorer OS for stage IV versus stages I/II (P = .15; Fig 3; Table 6). Additionally, among patients with CNS disease at diagnosis, there was a trend for a poorer OS (P = .1) but not EFS (P = .3) in patients with an LDH serum level greater than 500 U/L as compared with other cases.
Overall, 8.1% of patients with newly diagnosed NHL at our institution had CNS involvement, a frequency higher than the 4% reported in the Berlin-Frankfurt-Munster (BFM) and Children’s Cancer Group (CCG) trials (Table 7).13,14 Differences in criteria for defining CNS disease may account for some of these differences. For example, in the BFM 86 regimen, isolated CNP was not a defining feature for CNS disease.13 The rate of CNS involvement varies with disease sites at diagnosis and with histologic subtype. In our population, it was highest among patients with bone marrow and/or bone involvement. Interestingly, it was not significantly increased in children with head and neck involvement, who have generally been thought to be at increased risk for CNS disease. The reasons for this discrepancy are not clear. CNS involvement was also more frequent among those with SNCC NHL/B-cell ALL histology (13%). This frequency was not unlike the 8.4% reported by French investigators among a similar group of patients15 but was higher than the 4.0% and 4.4% reported by the CCG14 and BFM groups.13 Among our patients with lymphoblastic NHL, the frequency of CNS disease was intermediate to that seen in SNCC and large-cell cases and was consistent with that reported in other studies.13,14,16 CNS disease was least frequent in the large-cell subtype in our population. Of 119 children in this subgroup, only one presented with CNS disease evidenced by an orbital primary tumor and an associated CNP. In the BFM experience, none of the 53 children with large-cell NHL had CNS involvement.13 However, for reasons that are not clear, the CCG reported a 5.6% incidence among this subset of patients.14
CNS involvement is characterized by lymphoma cells in the CSF or by CNP on physical examination.2 Most of our patients had lymphoma cells in the CSF (64%), whereas 25% had CNP, usually involving either the facial nerve or those that innervate the extraocular muscles. A small percentage of patients presented with both features. There was no apparent difference in treatment outcome between those with CNP at diagnosis and those who presented with malignant CSF pleocytosis. The CNS was the most frequent site of relapse among children who had presented with CNS disease; the bone marrow was the second most frequent site of relapse. Of the nine children with CNS relapses, seven had malignant CSF pleocytosis and two had both pleocytosis and CNP. There was no obvious relationship between the pattern of presentation of CNS disease (CSF+ v CNP+) at relapse and at initial diagnosis. Although children who present with CNS disease are considered to be at higher risk of treatment failure, the factors that create a higher risk have not been extensively studied. In fact, some studies have questioned the prognostic significance of CNS disease at diagnosis.7,17,18 Of interest, in our review, involvement of the CNS at diagnosis did not have independent prognostic significance with respect to either EFS or OS. A CCG study of children with SNCC NHL suggested that the presence of CNS disease at diagnosis did not clearly predict an unfavorable outcome independent of initial bone marrow status.18 In a study by Haddy et al17 of adults and children with either newly diagnosed or relapsed SNCC NHL, the tumor burden, as reflected by serum LDH, seemed to be a more reliable predictor of outcome than CNS disease per se. This conclusion is supported by the findings of our study, which, in contrast to that of Haddy et al, deals solely with children newly diagnosed with NHL and associated CNS disease. Even among those children with CNS disease at diagnosis, we observed that tumor burden, whether reflected by serum LDH or by the disease stage irrespective of CNS involvement, was greater in patients who had a poor treatment outcome. These observations suggest that improvements in treatment outcome for children with NHL and CNS disease at diagnosis have and will continue to require advances in systemic therapy. In this regard, a recent POG study attributed the lack of prognostic significance of CNS disease at diagnosis in children with SNCC NHL to improved systemic therapy.7 Current investigational approaches to the treatment of CNS disease vary with NHL histologic type.19 In current POG and St Jude protocols, IT methotrexate and cranial irradiation are used to treat large-cell NHL with associated CNS disease. However, the relatively small number of these cases precludes a meaningful comparison of historical treatment strategies (eg, IT chemotherapy ± cranial radiation ± high-dose systemic therapy). In contrast, there is more experience with the treatment of CNS disease associated with the lymphoblastic and SNCC histologic types. Most current regimens incorporate multiagent IT chemotherapy coupled with cranial radiation for CNS+ lymphoblastic NHL cases. The approach to the treatment of CNS disease in cases of SNCC NHL, however, has been less uniform. Although most strategies incorporate high-dose systemic chemotherapy coupled with aggressive multiagent IT chemotherapy, the role of cranial radiation remains controversial. Although the very effective French LMB 86 and 89 regimens included cranial radiation for these cases, most protocols, including the current French Society of Pediatric Oncology regimen, have not.20,21 In most CNS+ SNCC cases, drugs have been delivered to the CNS via the IT route by lumbar puncture; however, we and investigators of the BFM group have used an intraventricular access device for intraventricular access device for drug delivery.13,22,23 This strategy for drug delivery may have contributed to the excellent result we observed using the POG 8617 regimen for children with CNS disease at diagnosis. The optimal approach to delivery of CSF chemotherapy has yet to be established; however, the technical difficulties that may be encountered with frequent lumbar punctures are generally avoided with the Ommaya reservoir. Our current management of CNS disease at diagnosis in children with SNCC NHL combines the Ommaya reservoir strategy for intraventricular drug delivery with a systemic therapy approach similar to that of LMB 89.24 As further information emerges about the relationship of treatment failure to histologic type and CNS disease at presentation, strategies for improving both systemic and CNS-directed therapy will become clearer.
Supported by grants CA-20180 and CA-21765 from the National Cancer Institute and by the American Lebanese Syrian Associated Charities. We thank Annette Stone and Mary Heim for data management, Wren Kennedy for chart review and helpful comments, Sharon Naron for editorial consultation, and Peggy Vandiveer for help in preparation of this manuscript.
1. Magrath IT: Malignant non-Hodgkin’s lymphomas in children, in Pizzo PA, Poplack DG (eds): Principles and Practice of Pediatric Oncology, 2nd ed. Philadelphia, PA,JB Lippincott, 1993 pp 536–575 2. Murphy SB: Classification, staging and end results of treatment of childhood non-Hodgkin’s lymphoma: Dissimilarities from lymphoma in adults. Semin Oncol 7:332–339, 1980[Medline] 3. National Cancer Institute: National Cancer Institute sponsored study of classifications on non-Hodgkin’s lymphomas: Summary and description of a working formulation for clinical usage—The Non-Hodgkin’s Lymphoma Pathologic Classification Project. Cancer 49:2112–2135, 1982[Medline] 4. Murphy S, Fairclough DL, Hutchison RE, et al: Non-Hodgkin’s lymphomas of childhood: An analysis of the histology, staging, and response to treatment of 338 cases at a single institution. J Clin Oncol 7:186–193, 1989[Abstract] 5. Murphy SB, Hustu HO: A randomized trial of combined modality therapy of childhood Hodgkin’s lymphoma. Cancer 45:630–637, 1980[Medline] 6. Murphy SB, Bowman WP, Abromowitch M, et al: Results of treatment of advanced-stage Burkitt’s lymphoma and B cell (Sig+) acute lymphoblastic leukemia with high-dose fractionated cyclophosphamide and coordinated high-dose methotrexate and cytarabine. J Clin Oncol 4:1732–1739, 1986[Abstract]
7.
Bowman WP, Shuster JJ, Cook B, et al: Improved survival for children with B-cell acute lymphoblastic leukemia and stage IV small noncleaved-cell lymphoma: A Pediatric Oncology Group study. J Clin Oncol 14:1252–1261, 1996 8. Agresti A: Categorical Data Analysis. New York, NY,John Wiley & Sons, 1990 9. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481, 1958 10. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomized clinical trials requiring prolonged observation of each patient: II. Analysis and examples. Br J Cancer 35:1–39, 1977[Medline] 11. Mantel N. Haenszel W: Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 22:719–748, 1958 12. Cox DR: Regression models and life tables (with discussion). J Roy Stat Soc Ser B 34:187–220, 1972
13.
Reiter A, Schrappe M, Parwaresch R, et al: Non-Hodgkin’s lymphomas of childhood and adolescents: Results of a treatment stratified for biologic subtypes and stage—A report of the Berlin-Frankfurt-Münster Group. J Clin Oncol 13:359–372, 1995
14.
Anderson JR, Derek R, Jenkin DT, et al: Long-term follow-up of patients treated with COMP or LSA2L2 therapy for childhood non-Hodgkin’s lymphoma: A report of CCG-551 from the Children’s Cancer Group. J Clin Oncol 11:1024–1032, 1993
15.
Patte C, Philip T, Rodary C, et al: Improved survival rate in children with Stage III and IV B cell non-Hodgkin’s lymphoma and leukemia using multi-agent chemotherapy: Results of a study of 114 children from the French Pediatric Oncology Society. J Clin Oncol 4:1219–1226, 1986 16. Patte C, Kalifa C, Flamant F, et al: Results of the LMT81 protocol, a modified LSA2L2 protocol with high dose methotrexate, on 84 children with non-B-cell (lymphoblastic) lymphoma. Med Pediatr Oncol 20:105–113, 1992[Medline]
17.
Haddy TB, Adde MA, Magrath IT: CNS involvement in small noncleaved-cell lymphoma: Is CNS disease per se a poor prognostic sign? J Clin Oncol 9:1973–1982, 1991
18.
Gururangan S, Sposto R, Meadows A, et al: Outcome of central nevous system (CNS) disease at diagnosis of small non-cleaved cell lymphoma (SNCCL) is dependent on marrow status. J Clin Oncol 18:2017–2025, 2000
19.
Sandlund JT, Downing JR, Crist WM: Non-Hodgkin’s lymphoma in childhood. N Engl J Med 334:1238–1248, 1996 20. Patte C, Leverger G, Michon J, et al: High survival rate of childhood B-cell lymphoma and leukemia (ALL) as result of the LMB89 protocol of the SFOP: Proceedings of the Fifth International Conference on Malignant Lymphoma, June 9–12, 1993, Lugano, Switzerland 21. Patte C, Leverger G, Perel Y, et al: Results of LMB 86 and 89 protocols of the French Pediatric Oncology Society (S.F.O.P.) for children with B-cell lymphoma and leukemia with CNS involvement (CNS+). Med Pediatr Oncol 21a:535, 1993 (abstr)
22.
Reiter A, Schrappe M, Ludwig W: Favorable outcome of B-cell acute lymphoblastic leukemia in childhood: A report of three consecutive studies of the BFM Group. Blood 80:2471–2478, 1992 23. Strother DR, Glynn-Barnhart A, Kovnar E, et al: Variability in the disposition of intraventricular methotrexate: A proposal for rational dosing. J Clin Oncol 11:1741–1747, 1989 24. Patte C, Michon J, Behrendt H, et al: Results of the LMB 89 protocol for childhood B-cell lymphoma and leukemia (ALL): Study of the SFOP (French Pediatric Oncology Society). Med Pediatr Oncol 29:358, 1997 (abstr) This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2000 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 test.
500 U/L) and stage influenced both OS and EFS (P = .0002 and 0.016, respectively, and P = .0001 and P = .0001, respectively). The relationship between CNS disease and serum LDH at diagnosis is depicted in 
