- © 2002 by American Society of Clinical Oncology
Non-Hodgkin’s Lymphoma Arising in Bone in Children and Adolescents Is Associated With an Excellent Outcome: A Children’s Cancer Group Report
- Mark A. Lones,
- Sherrie L. Perkins,
- Richard Sposto,
- Nicole Tedeschi,
- Marshall E. Kadin,
- Carl R. Kjeldsberg,
- John F. Wilson,
- David L. Zwick and
- Mitchell S. Cairo
- From the Pathology Department, Children’s Hospital of Orange County/St Joseph Hospital, Orange, CA; Department of Pathology, University of Utah Health Sciences Center, Salt Lake City, UT; Operations Center, Children’s Oncology Group, Arcadia, CA; Keck School of Medicine, University of Southern California, Los Angeles, CA; Department of Pathology, Beth Israel Deaconess Hospital and Harvard Medical School, Boston, MA; Pathology Department, Children’s Mercy Hospital, Kansas City, MO; and Division of Pediatric Oncology, Children’s Hospital of New York, Columbia University, New York, NY.
- Address reprint requests to Mark A. Lones, MD, Children’s Oncology Group, PO Box 60012, Arcadia, CA 91066-6012; email: smason@ childrensoncologygroup.org.
Abstract
PURPOSE: Non-Hodgkin’s lymphoma (NHL) arising in bone is a heterogeneous histologic type of NHL that includes large-cell lymphoma, lymphoblastic lymphoma, and small noncleaved-cell lymphoma. NHL arising in bone is well recognized in adults but is less well characterized and infrequent in children and adolescents.
PATIENTS AND METHODS: We performed a retrospective review of Children’s Cancer Group (CCG) studies treating children and adolescents with NHL over a 20-year period (CCG-551, CCG-501, CCG-502, CCG-503, CCG-552, CCG-5911, and CCG-5941) and determined the response and event-free survival (EFS) rates in 31 patients with NHL arising in bone.
RESULTS: The patients ranged in age from 3 to 17 years (median, 11 years; mean, 11 years), and 64.5% were male. All 31 (100%) patients achieved complete response. For 31 patients with NHL arising in bone, the product-limit estimated 5-year EFS was 83.8% ± 6.7%. EFS in 17 patients with localized disease (Murphy stages I and II) was 94.1% ± 5.7%, and EFS in 14 patients with disseminated disease (Murphy stage III) was 70.7% ± 12.4% (log-rank P = .10). EFS in 17 patients treated with chemotherapy and radiation was 70.1% ± 11.2%, and EFS in 14 patients treated with chemotherapy without radiation was 100% (P = .03). EFS in 26 patients with histology-directed treatment (LSA2-L2 or ADCOMP for lymphoblastic, other therapy for nonlymphoblastic) was 92.2% ± 5.3%, and in five patients with nonhistology-directed treatment it was 40.0% ± 21.9% (P < .001).
CONCLUSION: NHL arising in bone is a heterogeneous type of NHL that makes up approximately 2.0% of NHL in children and adolescents on CCG studies. Response and survival in this young age group seem superb, with histology-directed treatment protocols without radiation in both localized and disseminated disease.
NON-HODGKIN’S lymphoma (NHL) arising in bone as the primary site is unusual, comprising less than 5% of all NHLs and 7% of primary bone tumors.1 The peak incidence is in the fifth decade of life, with a predominance of male patients and most lesions occurring in the lower half of the body.2,3 Approximately half of NHLs arising in bone occur in the long bones, with pelvic and vertebral lesions making up another 25% of cases.1 Clinical presentations include localized bone pain and occasionally a palpable mass. Radiological studies will usually demonstrate an osteolytic lesion.1,3
Pathologic classification of NHL arising in bone has demonstrated a wide spectrum of histologic subtypes, although most are classified as intermediate-grade or aggressive B-cell lymphomas.2-6 Historically, the most common histologic subtype is diffuse large B-cell lymphoma,1,3,4 although other histologic subtypes including CD30-positive anaplastic large-cell lymphomas (predominantly of T-cell type),7,8 Burkitt lymphomas,3,5 and lymphoblastic lymphomas (including both precursor B-cell and precursor T-cell phenotypes)9 have also been described.
Although several studies have been undertaken in adults or patient cohorts of mixed ages,2-5,10-20 NHL arising in bone in children and adolescents has been less well studied, and large studies with long-term clinical follow-up are few.7,21-26 We reviewed 31 cases of NHL arising in bone in children and adolescents derived from the Children’s Cancer Group (CCG) experience, with emphasis on clinical outcome and approach to clinical staging, and have compared our results with the present literature applicable to this subject.
PATIENTS AND METHODS
Patient Selection
A retrospective review of CCG NHL studies was undertaken to identify patients with NHL arising in bone. These studies covered the time period from 1977 to 1996 and included the following protocols: CCG-551, Non-Hodgkin’s Lymphoma (1977 to 1982)27-29; CCG-501, Localized Non-Hodgkin’s Lymphoma (1982 to 1986)30; CCG-502, Lymphoblastic Lymphoma (1983 to 1990)31; CCG-503, Non-Lymphoblastic Non-Hodgkin’s Lymphoma, Disseminated (1983 to 1990)32; CCG-552, Disseminated Non-Lymphoblastic Lymphoma of Childhood (1986 to 1989)33; CCG-5911, A Pilot Study in the Treatment of Childhood Non-Localized Non-Lymphoblastic Non-Hodgkin’s Lymphoma (1991 to 1993)34; and CCG-5941, A Pilot Study for the Treatment of Newly-Diagnosed Disseminated Lymphoblastic Lymphoma (1994 to 1996).35 Biopsies from patients with NHL from these protocols were confirmed by central pathology review.35-37 This study included patients with NHL restricted to disease involvement of bone, including direct extension into adjacent tissues. Patients with NHL involving bone and other sites (lymph nodes, organs, CSF, and bone marrow) were excluded.
Staging Evaluation
The extent of disease was determined by clinical staging methods, which varied among the protocols. Radiology studies used included chest x-ray, computed tomography (CT) scan, bone scan, liver/spleen scan, gallium scan, intravenous pyelogram, and abdominal/pelvic ultrasound. Pathology evaluations included CSF examination and bone marrow examination (bone marrow aspirate only or aspirate with biopsy). All patients with NHL arising in bone enrolled on these protocols were classified as having disseminated disease and treated identically to those with disseminated disease on the basis of initial prognostic observations in CCG-551 and pediatric NHL staging schemes by the CCG.28 For this retrospective study, patients were reclassified for stage of disease using a modification of the Murphy stage.38
Pathology Evaluation
Biopsy tissue was processed and evaluated at the primary institution. Central pathology review during each respective CCG study consisted of light microscopic evaluation of morphology only, with details for some studies presented previously.37 Classification was performed according to the Rappaport classification39 or National Cancer Institute Working Formulation.40,41 There were no immunophenotyping studies performed on these biopsies during the central pathology review for these protocols.
Therapy
Because the CCG staging scheme for pediatric NHL classified NHL arising in bone as disseminated disease, no patients with NHL arising in bone were identified from the CCG-501 study of localized NHL, indicating that none of the patients in this retrospective study were treated with short-duration chemotherapy for localized NHL at other sites.30 No patients with NHL arising in bone were identified from the CCG-552 study of disseminated nonlymphoblastic NHL.33 CCG-552 was a limited-institution pilot study that enrolled only 68 patients and was limited to small noncleaved-cell or large-cell lymphomas that were disseminated according to Murphy stage.
Patients with NHL arising in bone were identified from the CCG-551, CCG-502, CCG-503, CCG-5911, and CCG-5941 studies. These patients were all treated with intense therapeutic regimens for disseminated NHL, which varied among the different protocols and were usually of long duration. The treatment regimens for NHL arising in bone included the following: COMP (cyclophosphamide, vincristine, methotrexate, prednisone) with radiation (CCG-551)27,28; modified LSA2-L2 (COMP with daunomycin, cytarabine, thioguanine, asparaginase, carmustine, hydroxyurea) with radiation (CCG-551, CCG-502)27,28,31; ADCOMP (asparaginase, daunomycin, COMP) with radiation (CCG-502)31; COMP (CCG-503)32; COMP+D (daunomycin, COMP) (CCG-503)32; Orange (CHOP with etoposide, ifosfamide, methotrexate, cytarabine, dexamethasone, l-asparaginase, cisplatin) (CCG-5911)34; French (cyclophosphamide, vincristine, prednisone, doxorubicin, methotrexate, cytarabine, and etoposide) (CCG-5911); and intensive NY1 (cyclophosphamide, vincristine, daunomycin, prednisone, asparaginase, cytarabine, methotrexate) (CCG-5941).35
Response Evaluation
Response to therapy was assessed at each institution after two general phases of therapy: postinduction and postconsolidation or maintenance. The responses were classified in the protocols as follows: complete response (CR), disappearance of all disease; partial response (PR), decrease of 50% or more of all measurable lesions; no response (NR), all lesions remain within 50% of initial size; or progressive disease (PD), any new lesion or increase of any measurable lesion by greater than 25% in size or new infiltration of the bone marrow or CNS.
Statistical Analysis
The primary end point for statistical survival analysis was event-free survival (EFS), defined as the minimum time to disease progression, disease relapse, occurrence of a second malignant neoplasm, or death from any cause. Secondary end points included overall survival (OS), which was the time from study entry to death from any cause. EFS and OS curves and probabilities were computed using the product-limit estimate with Greenwood SEs. Differences in long-term EFS were tested using the log-rank test.42 All quoted P values are two-sided.
RESULTS
Case Selection
A total of 1,584 cases of NHL were identified from previous CCG studies.27,28,30-36 Thirty-one (2%) patients were classified by clinical and pathologic features as having NHL arising in bone restricted to involvement of bone including direct extension into adjacent tissues. These 31 patients were identified from the CCG-551, CCG-502, CCG-503, CCG-5911, and CCG-5941 studies.
Demographics
The clinical features at presentation of the patients with NHL arising in bone are listed in Table 1. Twenty (65%) patients were male. The median age was 11 years (mean, 11 years; range, 3 to 17 years). For males, the median age was 13 years (mean, 11 years; range, 3 to 17 years). The 11 females had a median age of 8 years (mean, 10 years; range, 3 to 16 years). Eighteen (58%) patients were children (< 13 years of age), whereas 13 (42%) were adolescents (≥ 13 years of age). The distribution of race among these patients was as follows: white, 23; black, four; Hispanic, two, and other, two.
Clinical Features of NHL Arising in Bone in Children and Adolescents in CCG-551, CCG-502, CCG-503, CCG-5911, and CCG-5941 Trials
Clinical Evaluation
Clinical evaluations at the time of diagnosis, including radiology studies, demonstrated no disease at any separate or distant sites outside of bone. In 17 patients in whom skeletal surveys had been performed, lesions were demonstrated in all patients. In 28 patients evaluated by bone scan, results were positive in 93%. In 13 patients evaluated by gallium scan, results were positive in 69% (large-cell lymphoma in one of two, lymphoblastic lymphoma in seven of nine, and small noncleaved-cell lymphoma one of two). In all 31 patients, CSF and bone marrow examinations had no evidence of lymphoma.
Pathology Evaluation
The biopsies in this series were nearly all surgical biopsy specimens (one needle core biopsy). There were no cases in which the diagnostic material represented fine-needle aspirate material. The 31 patients with NHL arising in bone consisted of 14 with large-cell lymphoma, 14 with lymphoblastic lymphoma, and three with small noncleaved-cell lymphoma. There was no immunophenotypic evaluation of these biopsies.
Sites of Disease
In the patients with NHL arising in bone, there was a spectrum of sites of disease involvement (Table 1). Bone sites of involvement by NHL included femur in 14, humerus in nine, tibia in eight, vertebrae in seven, rib in six, skull in six, ilium in five, scapula in four, acetabulum in three, ischium in three, mandible in three, radius in two, fibula in one, maxilla in one, metatarsal in one, sacrum in one, sternum in one, and ulna in one patient.
Stage of Disease
The stage of disease for NHL arising in bone included a spectrum of single or multiple sites of bone involvement (Table 1). The Murphy stage of disease in these 31 patients was stage I in 13, stage II in four, and stage III in 14 patients. In the 14 patients with large-cell lymphoma, the Murphy stage of disease was as follows: stage I in six, stage II in three, and stage III in five patients. The Murphy stage of disease in the 14 patients with lymphoblastic lymphoma was as follows: stage I in six, stage II in zero, and stage III in eight patients. For the three patients with small noncleaved-cell lymphoma, the Murphy stage was as follows: stage I in one, stage II in one, and stage III in one patient.
Therapeutic Response
The treatment and outcome for patients with NHL arising in bone are listed in Table 1. All 31 patients with NHL arising in bone achieved CR. However, evaluation of response is difficult to determine, because osteolytic lesions on skeletal survey films do not demonstrate changes during the early treatment period. Patients were not uniformly evaluated and followed to evaluate response with bone scan or gallium scan. There were no patients with progressive disease.
Relapse
Relapse occurred in five (16.1%) patients at intervals ranging from 3 to 17 months after study entry (Table 1). These five patients were from 5 to 16 years of age at study entry. Relapse sites included new bone sites only in two patients (case 1, large-cell; case 19, lymphoblastic), bone marrow as acute leukemia in one patient (case 15, lymphoblastic), new bone site and bone marrow as acute leukemia in one patient (case 24, lymphoblastic), and primary site associated with peripheral blood blasts in one patient (case 29, small noncleaved-cell). There were no other extranodal or nodal sites of relapse reported. All five patients with relapse had previously received radiation therapy. All five patients had relapse sites outside of the radiation field, along with two patients with relapse sites within the radiation field (case 1, large-cell; case 29, small noncleaved-cell). All five patients with relapse of NHL progressed during subsequent therapy and later died of disease within an interval of 3 to 14 months after relapse.
Among the patients with relapse of disease, four occurred on CCG-551 and one occurred on CCG-502. Relapses occurred in four patients with disseminated disease (Murphy stage III) and one patient with localized disease (Murphy stage I). Three of the patients with relapse of disease were initially treated with chemotherapy regimens that are now considered less effective (case 1, large-cell lymphoma Murphy stage III treated with LSA2-L2 and radiation therapy; case 19, lymphoblastic lymphoma Murphy stage III treated with COMP and radiation therapy; and case 29, small noncleaved-cell lymphoma Murphy stage I treated with LSA2-L2 and radiation therapy). Two patients (cases 15 and 24) had lymphoblastic lymphoma Murphy stage III treated with LSA2-L2 and radiation therapy and subsequently relapsed as acute lymphoblastic leukemia, along with a new bone site of disease in case 24.
Other Events
There were no reports of second malignancy associated with radiation therapy in the 17 patients who received radiation therapy, of which 10 are survivors at longer than 5 years.
Survival in NHL Arising in Bone
Overall, 26 patients are alive at last contact (Table 1). For the group of 31 patients with NHL arising in bone, the product-limit estimated 5-year EFS was 83.8% ± 6.7% (Fig 1), and OS was 83.5% ± 6.7%. The median follow-up interval was 9 years (96% > 3 years, 85% > 5 years) in the patients without disease recurrence or death.
Fig 1. Product-limit estimated 5-year EFS for all 31 patients, 17 patients with localized (Murphy stage I and II), and 14 patients with disseminated (Murphy stage III) NHL arising in bone in children and adolescents.
Survival results for subgroups of the 31 patients with NHL arising in bone are described below. The 5-year EFS for localized disease (Murphy stages I and II) was excellent at 94.1% ± 5.7%, whereas disseminated disease (Murphy stage III) was less favorable at 70.7% ± 12.4% (Fig 1). The analysis of difference in EFS between the localized and disseminated disease patients yielded a P value of .10. For histologic types of NHL arising in bone, 5-year EFS was excellent in the 14 patients with large-cell lymphoma at 92.8% ± 6.9%, whereas it was less favorable in the 14 patients with lymphoblastic lymphoma, at 77.9% ± 11.3%, and small noncleaved-cell lymphoma consisted of only three patients, at 66.7% ± 27.2%. When classified by histologic type and stage of disease, the nine patients with localized large-cell lymphoma (Murphy stages I and II) and the six patients with localized lymphoblastic lymphoma (Murphy stage I) had an excellent 5-year EFS of 100%, whereas the patients with disseminated disease (Murphy stage III) seemed to have a less-favorable outcome (five disseminated large-cell, 80.0% ± 17.9%; eight disseminated lymphoblastic, 60.0% ± 18.2%).
For the 17 patients treated with combination chemotherapy and radiation therapy, the 5-year EFS was 70.1% ± 11.2%, whereas the 14 patients treated with chemotherapy without radiation therapy had a 5-year EFS of 100% (P = .03), indicating that EFS in patients with NHL arising in bone treated with chemotherapy and radiation therapy may be worse than in the group treated with chemotherapy without radiation therapy. However, the assignment to radiation therapy was not randomized but was determined by the era of treatment. Also, assignment to chemotherapy regimen was not based on histology during CCG-551.
For patients treated with histology-directed therapy (LSA2-L2 or ADCOMP for lymphoblastic lymphoma; other therapy for nonlymphoblastic lymphoma, large-cell or small noncleaved-cell), the 5-year EFS was 92.2% ± 5.3% (Fig 2), whereas for patients treated with nonhistology-directed therapy, the 5-year EFS was 40.0% ± 21.9% (P < .001). Hence, EFS in patients with NHL arising in bone treated with histology-directed chemotherapy seems superb and better than in the group treated with nonhistology-directed chemotherapy regimens.
Fig 2. Product-limit estimated 5-year EFS for 26 patients treated with histology-directed chemotherapy (LSA2-L2 or ADCOMP for lymphoblastic; other therapy for nonlymphoblastic, large-cell or small noncleaved-cell) and five patients treated with nonhistology-directed chemotherapy regimens with NHL arising in bone in children and adolescents (log-rank P < .001).
Among patients treated with histology-directed therapy, the 14 patients with localized disease (Murphy stages I and II) had an excellent outcome, with 5-year EFS at 100%, and the 12 patients with disseminated disease (Murphy stage III) also had a favorable outcome at 82.5% ± 11.3% (P = .11).
DISCUSSION
Although NHL arising in bone has been described predominantly as occurring in adults,1-5,7,8,10-20 it also occurs in children and adolescents. Reports of survival in adults have been variable, with overall survival rates ranging from 40% to 90% and therapy often including radiation. NHL arising in bone is a heterogeneous type of NHL of B-cell or T-cell origin.
In children and adolescents, there have been a few studies with only small numbers of patients with NHL arising in bone and limited documentation of phenotype.21-26 NHL arising in bone seems to be of low frequency (1% to 2%) on the basis of studies from single institutions containing the largest numbers of NHL patients.24,25 These studies in young patients, which contain heterogeneous types of NHL, have reported variable survival rates (Table 2) ranging from 40% to 100%. The previous studies and present study include a total of 87 cases, which are predominantly localized disease (Murphy stages I and II, 63 cases [72%]; Murphy stage III, 24 cases [28%]). Comparison of results among these studies is limited because of the variable treatment regimens used even within single reports. There seems to be an increased EFS among children and adolescents with localized disease (Murphy stages I and II) in the range from 75% to 100% compared with disseminated disease (Murphy stage III) ranging from 25% to 71%.
Comparison of Reports on NHL Arising in Bone in Children and Adolescents
The earlier studies in children and adolescents21-25 included few patients with NHL that was restricted to bone and histologically classifiable (Tables 2 and 3⇓). These patients included those with NHLs restricted to bone with either localized or disseminated Murphy stage (Table 2), representing a variety of histologic types of NHL (Table 3). Among these early studies and the present study, the histologic types of NHL are composed predominantly of large-cell lymphomas, along with a substantial proportion of lymphoblastic lymphomas, whereas there are only few small noncleaved-cell lymphomas (Table 3). The large-cell lymphomas occur more often as localized disease, whereas the lymphoblastic lymphomas more frequently present as disseminated disease arising in bone (Table 3).
Comparison of Reports of Histologic Type and Stage of NHL Arising in Bone in Children and Adolescents*
In the early studies,21-25 patients with NHL arising in bone were treated with a variety of chemotherapy regimens for disseminated NHL, supplemented with radiation therapy in most cases. From these early studies of 29 patients, there was relapse of NHL in six patients,22,24,25 and two patients developed bone sarcomas within the radiation field.21 However, in the study by Coppes et al,25 it is important to recognize that the three failures occurred on treatment regimens that are presently considered less effective for the histologic types of NHL (one large-cell lymphoma treated with modified LSA2-L2 and two lymphoblastic lymphomas treated with COPAD25).
A more recent study from the Pediatric Oncology Group26 described 31 children and adolescents with localized NHL involving bone, including 27 patients reported as localized NHL restricted to bone (Table 2). The treatment regimens consisted of variations of CHOP chemotherapy, including short duration of therapy in some and radiation therapy in a small subset of patients. The entire group of 31 patients achieved CR, and 30 are without events. Twenty patients from this report were treated with short-duration CHOP chemotherapy, with encouraging results of 19 patients without events (one relapse). However, the follow-up interval for all patients from this study is short, with 16 patients followed for less than 3 years, which represents the most recent group with the short-duration CHOP chemotherapy regimen. For these 27 patients26 with localized NHL restricted to bone, the histologic type of NHL and corresponding specific treatment regimens and outcomes could not be determined from the report (Tables 2 and 3⇑).
The present study of 31 patients with NHL arising in bone treated on CCG treatment protocols demonstrates results similar to those on the previous studies combined (Tables 2 and 3⇑). Overall, these 31 patients had an excellent outcome. On the basis of the results of the present and previous studies listed in Table 2, there seems to be a lower survival rate in the group with disseminated disease (Murphy stage III) than in the group with localized disease (Murphy stages I and II), suggesting that extent of disease classified by a modified Murphy stage in NHL arising in bone is useful for correlation with outcome. The high survival rates in patients with NHL arising in bone that have localized disease (Murphy stages I and II) suggest that these patients should not be classified as poor prognosis or disseminated disease in staging classifications for NHL in children and adolescents, as done previously in the CCG. The CCG has previously reported excellent survival in patients with localized NHL at other sites treated with short-duration chemotherapy.29,30 In addition, there seems to be excellent survival in localized NHL arising in bone in the recent Pediatric Oncology Group study26 using short-duration chemotherapy. In addition, failures in the present and previous studies occurred frequently in patients with disseminated disease (Murphy stage III) and chemotherapy regimens that are presently considered less effective for the histologic type of NHL.27-29 Although there are only few patients treated with a variety of chemotherapy regimens, the lower survival rates in patients with disseminated disease suggest that NHL arising in bone in children and adolescents should be treated with chemotherapy regimens directed to histologic type of NHL. This is additionally supported in the present study by the small group of patients with disseminated NHL arising in bone treated with histology-directed therapy, which seems to have a more favorable survival rate, approaching the excellent survival rate of patients with localized NHL arising in bone.
Radiation therapy has been associated with increased risk of secondary malignancies in the radiation field, such as sarcomas in two patients with NHL arising in bone.21 Although this retrospective study of NHL arising in bone from CCG trials is small, the excellent survival rate (100%) in patients managed without radiation therapy indicates that chemotherapy alone is likely adequate for this disease. The recent study by Suryanarayan et al26 on localized NHL arising in bone also indicates that radiation therapy is not required.
Monitoring response to therapy is difficult in NHL arising in bone. Skeletal survey films do not show healing of lytic lesions early during therapy. Bone scans do not return to normal for a prolonged time period because of regenerating bone.24,25 Gallium scan is not positive in many patients, as revealed in this retrospective study. Presently, the most reliable method to document complete remission may be surgical biopsy,25 but this procedure may be limited by sampling error or complications such as fracture. However, positron emission tomography scans may be a more useful method than other radiologic techniques to evaluate response in future studies.43
NHL arising in bone is composed of different histologic types of lymphoma. In the present study, there are no immunophenotypic data available for these cases, because these studies were not performed as part of the central pathology review. Previous studies on NHL arising in bone in children and adolescents have included only limited or no data on immunophenotype.21-26 Future pathologic studies of NHL arising in bone in children and adolescents should include histologic and immunophenotypic examination to determine specific types of NHL (ie, diffuse large B-cell, T-cell anaplastic large-cell, precursor B or T lymphoblastic, or Burkitt).44,45
In conclusion, NHL arising in bone occurs in children and adolescents in low frequency, representing 2% of all patients with NHL on CCG studies. For NHL arising in bone treated with chemotherapy with or without radiation, survival is excellent in localized disease, whereas disseminated disease seems less favorable. The disseminated NHLs arising in bone should be treated with chemotherapy regimens directed to histologic type of NHL. NHLs arising in bone probably do not require radiation therapy. Future therapeutic studies of NHL arising in bone should be directed toward specific types of NHL classified by histologic and immunophenotypic features.
APPENDIX
The appendix listing the participating principal investigators is available online at www.jco.org.
Participating Principal Investigators: Children’s Cancer Group
Acknowledgments
Supported by the Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Department of Health and Human Services, Bethesda, MD. Contributing Children’s Cancer Group investigators, institutions, and grant numbers are given in the Appendix, which is available online at www.jco.org.
ACKNOWLEDGMENT
We thank Linda Rahl, Shaun Mason, and Lucia Noll for the editorial assistance in the preparation of this manuscript.
Footnotes
-
Presented in part at the Forty-Second Annual Meeting of the American Society of Hematology, December 1-5, 2000, San Francisco, CA.
- Received June 4, 2001.
- Accepted February 11, 2002.
