2030 Clinical Significance of Philadelphia Chromosome Positive Pediatric Acute Lymphoblastic Leukemia in the Context of Contemporary Intensive Therapies A Report from the Children’s Cancer Group Fatih M. Uckun, M.D., Ph.D.1 James B. Nachman, M.D.2 Harland N. Sather, Ph.D.3,4 Martha G. Sensel, Ph.D.4 Peter Kraft, M.S.5 Peter G. Steinherz, M.D.6 Beverly Lange, M.D.7 Raymond Hutchinson, M.D.8 Gregory H. Reaman, M.D.9 Paul S. Gaynon, M.D.10 Nyla A. Heerema, Ph.D.11 1 Children’s Cancer Group ALL Biology Reference Laboratory and Hughes Institute, St. Paul, Minnesota. 2 Section of Pediatric Hematology-Oncology, University of Chicago, Chicago, Illinois. 3 Department of Preventive Medicine, University of Southern California, Los Angeles, California. 4 Group Operations Center, Children’s Cancer Group, Arcadia, California. 5 Division of Biometry, University of Southern California, Los Angeles, California. 6 Department of Pediatrics, Memorial Sloan-Kettering Cancer Center, New York, New York. 7 Division of Oncology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania. 8 Department of Pediatric Hematology-Oncology, University of Michigan, Ann Arbor, Michigan. BACKGROUND. Children with Philadelphia (Ph) chromosome positive (⫹) acute lymphoblastic leukemia (ALL) represent a subgroup at very high risk for treatment failure. In this report, the authors assessed the outcome of Ph⫹ ALL in a large cohort of children treated on contemporary intensive chemotherapy protocols of the Children’s Cancer Group (CCG). METHODS. This study included 1322 children enrolled between 1988 –1995 on CCG risk-adjusted studies for ALL who had centrally reviewed cytogenetic data. Thirty patients had a t(9;22)(q34;q11) translocation and were referred to as Ph⫹; 1292 were Ph negative(⫺). Outcome analyses used standard life table methods. RESULTS. Compared with Ph⫺ ALL patients, Ph⫹ ALL patients were more likely to be black (P ⫽ 0.008), age ⬎10 years (P ⫽ 0.02), and have a leukocyte count ⱖ50,000/L (P ⬍ 0.0001). Nearly all Ph⫹ (96.7%) and Ph⫺ (98.3%) patients achieved remission after induction therapy, yet event free survival outcome was significantly worse for Ph⫹ patients compared with Ph⫺ patients, with 4-year estimates of 20.1% (standard deviation [SD] ⫽ 9.1%) and 75.8% (SD ⫽ 1.2%), respectively (P ⬍ 0.0001). This difference was maintained among patients regardless of presenting leukocyte count, age, or early response to therapy. Ten Ph⫹ patients underwent bone marrow transplantation (BMT) at the time of first remission; six of these patients remained event free at the time of analysis, and represent the majority (six of eight) of patients surviving event free. CONCLUSIONS. The findings of the current study confirm that Ph chromosome positivity represents a significant independent adverse risk factor for childhood ALL that has not been abrogated by current intensive chemotherapy programs. BMT at the time of first remission, as well as other alternative strategies employing biotherapeutic agents, should be considered in future front-line trials for pediatric patients with Ph⫹ ALL. Cancer 1998;83:2030 –9. © 1998 American Cancer Society. KEYWORDS: Philadelphia chromosome, children, acute lymphoblastic leukemia, cytogenetics. 9 Department of Hematology-Oncology, Children’s National Medical Center and the George Washington University, Washington, D.C. 10 Department of Pediatric Hematology-Oncology, University of Wisconsin, Madison, Wisconsin. 11 Department of Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, Indiana. Supported in part by research grants, including Children’s Cancer Group Chairman’s grant CA-13539, CA-51425 and CA-60437 from the Na© 1998 American Cancer Society tional Cancer Institute, National Institutes of Health. Dr. F. M. Uckun is a Stohlman Scholar of the Leukemia Society of America. Address for reprints; Fatih M. Uckun, M.D., Ph.D., Children’s Cancer Group ALL Biology Reference Laboratory and Hughes Institute, 2665 Long Lake Road, Suite 330, St. Paul, MN 55113. Received February 4, 1998; revision received May 20, 1998; accepted June 5, 1998. Ph Chromosome Positive Childhood ALL/Uckun et al. T he Philadelphia (Ph) chromosome, a cytogenetic abnormality that is characterized by deletion or translocation of 22q11, was described first in cases of chronic myelogenous leukemia (CML)1,2 but is now known to occur in 2–5% of all cases of childhood and in 16 –25% of adult acute lymphoblastic leukemia (ALL).2–5 In ALL, the Ph chromosome arises most often through a t(9;22)(q34;q11), with translocation of the c-abl proto-oncogene on chromosome 9 to either of the two breakpoint regions within the BCR gene on chromosome 22, which results in production of either a 210-kDa or a 190-kDa fusion protein with tyrosine kinase activity.6 – 8 Presence of a Ph chromosome appears to be an independent adverse risk factor for children with ALL3–5 despite the fact that it is associated with unfavorable presenting features, such as high leukocyte count, older age, and French–American–British (FAB) L2 morphology. Recent studies have suggested that small subsets of Ph positive (Ph⫹) patients who have low leukocyte count at diagnosis,9 show a “prednisone good response,”10 or are treated with reinforced early therapy followed by rotational therapy with noncross-resistant agents11 may have improved outcomes. The majority of children with Ph⫹ ALL, however, continue to experience extremely poor outcomes despite treatment on contemporary intensive treatment regimens that are effective for most other children with ALL. In the current report, we have assessed the incidence, prognostic significance, and treatment outcome of Ph⫹ ALL in a large cohort of children enrolled on risk-adjusted, contemporary, intensive treatment protocols of the Children’s Cancer Group (CCG). Our findings confirm that Ph chromosome positivity represents a significant adverse risk factor for ALL that has not been negated by current intensive chemotherapy regimens and that bone marrow transplantation (BMT) appears to represent a promising treatment modality for such patients. MATERIALS AND METHODS Patients Diagnosis of ALL was based on morphologic, biochemical, and immunologic features of the leukemic cells, including lymphoblast morphology on Wright– Giemsa-stained bone marrow smears, positive nuclear staining for terminal deoxynucleotidyl transferase (TdT), negative staining for myeloperoxidase, and cell surface expression of two or more lymphoid differentiation antigens, as described below. Between 1988 and 1995, a total of 3937 children (⬍21 years of age) were entered on the CCG studies included in this analysis (see below). Among these, 1322 patients had centrally reviewed and accepted cytogenetic data: 30 2031 (2.3%) had t(9;22) and are referred to as Ph⫹, and 1292 (97.7%) were Ph negative (Ph⫺). On CCG risk-adjusted ALL protocols, National Cancer Institute (NCI) standard risk patients12 were assigned either to CCG 1881, a low risk protocol for children 2–9 years of age and leukocyte count ⬍10,000/L, or to CCG 1891, an intermediate risk protocol for children 2–9 years of age and leukocyte count 10,000 – 49,999/L or 1 year of age and leukocyte count ⬍50,000/L. Patients who would be classified as NCI poor risk12 were assigned to CCG 1883, a protocol for infants less than 1 year of age, CCG 1882, a high risk protocol for patients 1–9 years of age with leukocyte count ⱖ50,000/L of age ⱖ10 years; or CCG 1901, a high risk protocol for patients with lymphomatous features. Definition of organomegaly (moderate or marked enlargement) and lymphomatous features were as described by Steinherz et al.13 Each protocol was approved by the NCI and the Institutional Review Boards of the participating CCG-affiliated institutions. Informed consent was obtained from parents, patients, or both, as deemed appropriate, according to Department of Health and Human Services guidelines. From 1988 to mid-1993, no specific provisions were made for referring Ph⫹ patients to BMT centers. In mid-1993, however, a CCG protocol (CCG 1921) employing allogeneic BMT in first remission was initiated for treatment of high risk patients, including those who were Ph⫹. Three patients in the current analysis were transferred to CCG 1921. Cytogenetic Analysis Cytogenetic analysis of leukemic cells was performed at diagnosis prior to initiation of therapy by local institutions. The recommended procedure called for preparation of banded chromosomes from unstimulated peripheral blood or direct and 24-hour-cultured preparations of fresh bone marrow, as described previously.14 Chromosome abnormalities were designated by using the 1995 International System for Human Cytogenetics Nomenclature.15 Abnormal clones were defined as two or more metaphase cells with identical structural abnormalities or extra chromosomes or three or more metaphase cells with identical missing chromosomes. Diagnosis of a normal karyotype required complete analysis of a minimum of 20 banded metaphases from bone marrow only. At least two original karyotypes of each abnormal clone or, in the case of normal cytogenetics, normal cells were reviewed by at least two members of the CCG cytogenetics committee. The term “primary clone” was used to identify the simplest karyotype, and assignment of ploidy group and other groupings were based on the primary clone. Deletions of chromosome 22 or 22q 2032 CANCER November 1, 1998 / Volume 83 / Number 9 were not considered to be cases of a Philadelphia chromosome and were not included in this analysis. Immunophenotyping Immunophenotypic analysis of leukemic cells was performed for a subset of patients in the current analysis. Mononuclear cell fractions comprised primarily (ⱖ90%) of leukemic cells were isolated from pretreatment bone marrow aspirate samples by centrifugation on Ficoll–Hypaque gradients. Immunophenotyping was performed centrally in the CCG ALL Biology Reference Laboratory by indirect immunofluorescence and flow cytometry, using monoclonal antibodies reactive with the following differentiation antigens: CD2, CD3, CD5, CD7, CD10, CD19, CD24, and CD34, as previously described.16 Patients were classified as B lineage if ⱖ30% of their leukemic cells were positive for CD19 or CD24 and if ⬍30% of their leukemic cells were positive for CD2, CD5, or CD7. Patients were classified as T lineage if ⱖ30% of their leukemic cells were positive for CD2, CD5, or CD7 and if ⬍30% of their leukemic cells were positive for CD19 or CD24. Statistical Methods Data analysis used patient information current to November, 1996. Patients with or without t(9;22) translocations were compared with respect to various clinical, demographic, and laboratory features by using chi-square tests for homogeneity of proportions. In selected cases, we also compared B-lineage t(9;22)⫹ and t(9;22)⫺ patients according to recently published NCI risk-classification criteria. The cohort of patients with accepted cytogenetic data was also compared with concurrently enrolled patients with unaccepted cytogenetic data. Most of the outcome analyses used life table methods and associated statistics. The primary endpoint examined was event free survival (EFS) from entry on study. EFS events included induction failure (nonresponse to therapy or death during induction), leukemic relapse at any site, death during remission, or second malignant neoplasm, which ever occurred first. Patients who did not experience an event at the time of EFS analysis were censored at the time of their last contact. A secondary endpoint, disease free survival (DFS), was the time to the first occurrence of an event (relapse at any site, death in remission, or second malignant neoplasm) from the start of the consolidation phase (i.e., postinduction). Patients without an event were censored at the time of last contact. Life table estimates were calculated by the Kaplan–Meier (KM) procedure, and the standard deviation (SD) of the life table estimate was obtained by using Greenwood’s formula.17 To indicate precision, the KM estimate of EFS and its SD are provided at selected time points. An approximate 95% confidence interval can be obtained from the life table estimate ⫾1.96 SDs. Life table comparisons of EFS outcome patterns for patient groups generally used the log rank statistic.18,19 P values for life table comparisons are based on the patterns of outcome across the entire period of patient follow-up but also may be given at specific time points for comparative purposes. P values ⱕ0.05 are referred to as significant. RESULTS Presenting Features of Phⴙ Patients Patients with Ph⫹ ALL were similar to Ph⫺ patients with respect to many presenting characteristics, yet several important differences were noted (Table 1). Compared with Ph⫺ ALL patients, Ph⫹ patients were more likely to be black (P ⫽ 0.008), older than 10 years of age (P ⫽ 0.02), and more likely to have leukocyte counts ⱖ50,000/L (P ⬍ 0.0001) and splenomegaly (P ⬍ 0.001). Indeed, the majority of Ph⫹ patients were clustered in the combined group with age ⱖ10 years and leukocyte counts ⱖ50,000/L. Among Ph⫹ patients with immunophenotypic data, 25 of 26 had B-lineage ALL. Moreover, 90.9% of B-lineage Ph⫹ patients had the immature, B-progenitor immunophenotype CD10⫹CD19⫹CD34⫹ compared with only 63.4% of B-lineage Ph⫺ patients. The majority of Ph⫹ ALL patients (86.7%) were assigned to high risk CCG protocols due to their unfavorable clinical features; only four Ph⫹ patients were assigned to the low or intermediate risk CCG studies, and no Ph⫹ patients were infants. Within the B-lineage subset, 84.0% of Ph⫹ patients met the NCI poor risk criteria. Most Ph⫹ patients (70.0%) had pseudodiploid karyotypes and, notably, only one had ⬎50 chromosomes in the primary clone. Presenting features for the entire cohort of patients for whom cytogenetic data were accepted generally were similar to those of concurrently enrolled patients for whom cytogenetic data were unaccepted or unavailable. However, patients with accepted cytogenetics were less likely to be in the “other” race group (P ⬍ 0.0001), more likely to have elevated hemoglobin (P ⬍ 0.0001) and platelet (P ⫽ 0.007) levels, and more likely to have T-lineage ALL (P ⫽ 0.02). Karyotypes of Phⴙ Patients Karyotypes of the 30 Ph⫹ ALL patients are shown in Table 2. With the exception of patient 19, all patients in this series presented with a balanced t(9;22)(q34; q11). Patient 19 had a complex rearrangement involving chromosome arms 7p, 7q, 9q, 19p, and 22q, resulting in an apparent Ph chromosome. In nine cases (patients 1, 6, 8, 18, 20, 22, 23, 25, and 30), a balanced Ph Chromosome Positive Childhood ALL/Uckun et al. 2033 TABLE 1 Presenting Features of Children with Phⴙ and Phⴚ Acute Lymphoblastic Leukemia Phⴚ (n ⴝ 1292) Phⴙ (n ⴝ 30) Variable Category No. (%) No. % P value* Age (yr) ⬍1 1–9 ⱖ10 ⬍20 20–49 ⱖ50 Male Female White Black Other Yes No Normal Mod. enlargeda Markedly enlarged Normal Mod. enlarged Markedly enlarged Normal Mod. enlarged Markedly enlarged Absent Small Large 1.0–7.9 8.0–10.9 ⱖ11.0 1–49 50–149 ⱖ150 Yes No L1, L1/L2 L2/L1, L2 B lineage T lineage Standard Poor Low Intermediate High Lymphomatous Infant Normal (46) Hypoploid (⬍46) Pseudodiploid (46) Hyperdiploid (47–50) Hyperdiploid (⬎46) 56 919 317 756 190 346 715 577 1046 85 159 30 1261 630 608 50 580 611 101 655 536 101 1171 47 73 650 422 207 605 387 295 56 1227 1121 170 789 168 470 278 257 392 433 154 56 399 69 347 143 334 4.3 71.1 24.5 58.5 14.7 26.8 55.3 44.7 81.1 6.6 12.3 2.3 97.7 48.9 47.2 3.9 44.9 47.3 7.8 50.7 41.5 7.8 90.7 3.6 5.7 50.8 33.0 16.2 47.0 30.1 22.9 4.4 95.6 86.8 13.2 82.4 17.6 62.8 37.2 19.9 30.3 33.5 11.9 4.3 30.9 5.3 26.9 11.1 25.9 0 16 14 5 3 22 15 15 23 6 1 0 30 14 14 2 3 25 2 11 18 1 30 0 0 17 8 4 19 4 7 1 29 25 5 25 1 4 21 1 3 23 3 0 0 5 21 3 1 0.0 53.3 46.7 16.7 10.0 73.3 50.0 50.0 76.7 20.0 3.3 0.0 100.0 46.7 46.7 6.7 10.0 83.3 6.7 36.7 60.0 3.3 100.0 0.0 0.0 58.6 27.6 13.8 63.3 13.3 23.3 3.3 96.7 83.3 16.7 96.2 3.8 16.0 84.0 3.3 10.0 76.7 10.0 0.0 0.0 16.7 70.0 10.0 3.3 0.02 Leukocyte count (⫻109/liter) Sex Race Down’s syndrome Liver Spleen Lymph nodes Mediastinal mass Hemoglobin (g/dL) Platelets (⫻109/liter) CNS disease at diagnosis FAB morphology Immunophenotype NCI risk groupb CCG study group Chromosome number ⬍0.0001 0.69 0.008 0.82 0.74 ⬍0.001 0.12 0.22 0.71 0.11 0.78 0.77 0.12 ⬍0.0001 ⬍0.0001 ⬍0.0001 Ph⫹: Philadelphia chromosome positive; Ph⫺: Philadelphia chromosome negative; central nervous system: FAB: French–American–British; NCI: National Cancer Institute; CCG: Children’s Cancer group. a Degree of organomegaly and size of mediastinal mass were determined as described in Materials and Methods. b Patients with B lineage acute lymphoblastic leukemia only. * Global chi-square test for homogeneity. 2034 CANCER November 1, 1998 / Volume 83 / Number 9 TABLE 2 Karyotypes of Children with Phⴙ Acute Lymphoblastic Leukemia Patient Karyotypea 1 2b 3b 4b,c 5c 6 7a 8 9d 10b 11 12b 13b 14c 15b 16d 17b 18 19b 20 21b 22 23 24b 25 26c 27 28b,d 29 30 46,XX,t(9;22)(q34;q11) 45,XY,dic(7;9)(p11;p11),t(9;22)(q34;q11) 46,XX,ider(9)(q10)t(9;22)(q34;q11),der(22)t(9;22)(q34;q11) 47,XY,del(9)(p22),t(9;22)(q34;q11),⫹der(22)t(9;22)(q34;q11)/46,idem,⫹9,⫺del(9)(p22) 55,XX,⫹X,⫹2,⫹4,⫹6,t(9;22)(q34;q11.2),⫹14,⫹18,⫹21,⫹21,⫹der(22)t(9;22)/55,idem,dup(1)(pter3q21::q323q21::q253q32::q253qter) 46,XY,t(9;22)(q34;q11) 46,XX,t(9;22)(q34;q11)/47,idem,t(8;11)(p10;q10),⫹der(22)t(9;22) 46,XY,t(9;22)(q34;q11) 45,XX,⫺7,t(9;22)(q34;q11) 46,XX,der(9)del(9)(p?21)t(9;22)(q34;q11),der(22)t(9;22)(q34;q11)/45,idem,⫺15 46,XY,t(9;22)(q34;q11)/46,idem,add(19)(p13)/46,idem,der(19)del(19)(p11)add(19)(q13) 46,XY,del(9)(p13),t(9;22)(q34;q11) 46,XY,t(9;22)(q34;q11)/46,idem,i(9)(q10),del(11)(q13q23) 46,XX,t(3;15)(p21;q15),t(9;22)(q34;q11)/47,idem,⫹der(22)t(9;22) 46,XX,der(9)del(9)(p?21)t(9;22)(q34;q11),der(22)t(9;22)(q34;q11) 45,XY,⫺7/45,idem,t(9;22)(q34;q11) 45,XX,del(9)(p13p24),t(9;22)(q34;q11),⫺15,der(20)t(15;20)(q13;q13.3) 46,XX,t(9;22)(q34;q11.2) 46,XY,der(7)t(7;19)(p15;p13.1)t(7;22)(q11.2;q11),ider(9)(q10)t(7;9)(p15;q34),der(19)t(7;19)(q11.2;p13.1),der(22)t(9;22)(q34;q11) 46,XX,t(9;22)(q34;q11.2) 47,XY,⫹8,ider(9)(q10)t(9;22)(q34;q11),der(22)t(9;22)(q34;q11)/48,idem,⫹mar 46,XX,t(9;22)(q34;q11) 46,XX,t(9;22)(q34;q11.2) 46,XX,del(1)(q32q42),⫺9,t(9;22)(q34;q11),⫹mar 46,XY,t(9;22)(q34;q11) 46,XX,t(9;22)(q34;q11)/58,idem,⫹X,⫹4,⫹4,⫹6,⫹10,⫹14,⫹14,⫹17,⫹18,⫹18,⫹21,⫹der(22)t(9;22) 49,XY,⫹X,t(9;22)(q34;q11),⫹17,⫹mar 45,X,⫺Y,⫺7,del(9)(p22),t(9;22)(q34;q11.2),del(10)(q22q24),der(12)t(Y;12)(q12;p13),⫹mar 46,XY,der(1)t(1;1)(p36;q21),t(9;22)(q34;q11) 46,XY,t(9;22)(q34;q11) a Karyotypic nomenclature according to ISCN guidelines15; additional abnormalities are indicated. 9p del. c ⫹der(22)t(9;22). d Monosomy 7. b t(9;22)(q34;q11) was the sole abnormality. Four additional cases (patients 7, 11, 13, and 26) with multiple clones had a clone with t(9;22) as the sole abnormality. Five patients (4, 5, 7, 14, and 26) had a duplicate Ph chromosome, i.e., ⫹der(22)t(9;22); all of these patients had complex karyotypes with additional abnormalities. Monosomy 7 was observed in only three patients (9, 16, and 28); it is noteworthy that this abnormality was the primary clone in patient 16, because ten cells had monosomy 7 as the sole abnormality, and three cells also had a Ph chromosome. Deletions of the short arm of chromosome 9 were observed in 12 patients including the complex rearrangement of patient 19. Four of these patients (3, 13, 19, and 21) had an i(9q), three of the der(9)t(9;22) and one of the homolog. In 5 of the 12 cases, the der(9)t(9;22) also had deletion or loss of 9p; the other 7 cases had deletion or loss of the homolog. Only one patient (5) was high hyperdiploid (⬎50), although a second patient (26) had a high hy- perdiploid clone secondary to a pseudodiploid primary clone. Treatment Outcome EFS outcome for the entire cohort of patients for whom cytogenetic data were available resembled that of concurrently enrolled patients for whom cytogenetics data were not available (P ⫽ 0.66; data not shown). Thus, patients who were included in the analysis of EFS outcome were representative of all patients concurrently enrolled in CCG ALL treatment programs. Although the majority of both Ph⫹ (96.7%) and Ph⫺ (98.3%) patients achieved remission following induction therapy, EFS outcome was significantly worse for Ph⫹ patients compared with Ph⫺ patients, with 4-year estimates of 20.1% (SD ⫽ 9.1%) and 75.8% (SD ⫽ 1.2%), respectively (P ⬍ 0.0001; Fig. 1). Median follow-up for all event free survivors was 39 months (range, 24 – 61 months). Of the eight patients who sur- Ph Chromosome Positive Childhood ALL/Uckun et al. FIGURE 1. Event free survival for children with Philadelphia (Ph) chromosome positive (Ph⫹) and negative (Ph⫺) acute lymphoblastic leukemia. Percentages of 30 Ph⫹ and 1292 Ph⫺ patients surviving event free during follow-up. Numbers of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years were 30, 16, 15, 11, 3, and 1, respectively, for Ph⫹ and 1292, 1186, 1091, 923, 665, and 393, respectively, for Ph⫺. vived event free at the time of the analysis, six underwent BMT in first remission (see below), and two received only chemotherapy. Seven of these eight patients (including all six who underwent BMT) remain event free with additional follow-up of 100 – 400 days; the other patient, who received only chemotherapy, had a marrow relapse at 3.8 years. By comparison, overall 7-year EFS for the entire cohort of patients enrolled on CCG studies during this treatment era was 70.7% (SD ⫽ 0.9%). Overall survival, although more favorable than EFS, was also significantly worse for Ph⫹ patients compared with Ph⫺ patients, with 4-year estimates of 56% (SD ⫽ 9.2%) and 85.3% (SD ⫽ 1.0%), respectively (P ⫽ ⬍ 0.0001). Six Ph⫹ patients remain alive after an event and range in follow-up from 1 year to 3.3 years postrelapse. One of these six patients has had a second relapse. The difference in outcome observed for Ph⫹ versus Ph⫺ patients was maintained within the subgroup of patients (n ⫽ 456) assigned to the high risk CCG 1882 study, with 4-year EFS of 11.3% (SD ⫽ 9.8%) and 73.4% (SD ⫽ 2.3%) for Ph⫹ and Ph⫺ patients, respectively (P ⬍ 0.0001). Among the three Ph⫹ patients enrolled on CCG-1901 (study for ALL with lymphomatous features), only one patient experienced an event; however, one of the patients who remained event free 2035 underwent BMT in first remission (see below). There were three patients enrolled on CCG 1891 (intermediate risk ALL) and one patient enrolled on CCG 1881 (low risk ALL); three of these four patients had events, and the one event free survivor underwent BMT in first remission. Poor outcome was observed for Ph⫹ patients regardless of age at diagnosis, with 4-year EFS rates of 21.2% (SD ⫽ 12.1%) and 28.6% (SD ⫽ 24.1%) for Ph⫹ patients ages ⬍10 years and ⱖ10 years, respectively (P ⬍ 0.25; Fig. 2). Likewise, the 4-year EFS for Ph⫹ patients with leukocyte count ⱖ50,000/L was 17.9% (SD ⫽ 10.3%), and six of the 8 Ph⫹ patients with leukocyte counts ⬍50,000/L experienced an event, with an estimated median EFS of 3.4 years (range 0.7–5.6 years; Fig. 3). Five patients had leukocyte counts ⱕ25,000/L; three of these patients had events at 3.4 years, 5.6 years, and 11 months after diagnosis. Furthermore, for the subset of 23 Ph⫹ ALL patients with Day 7 marrow status data who achieved a remission to induction therapy, initial response to treatment did not affect outcome: four-year DFS rates were 22.2% (SD ⫽ 12.8%) and 27.3%) for rapid early responders (ⱕ25% marrow blasts on Day 7 of induction chemotherapy)20 and slow early responders (⬎25% marrow blasts on Day 7 of induction chemotherapy), respectively (Fig. 4). It is interesting to note that among all older patients with leukocyte counts ⱖ50,000/L, those with B-lineage ALL had worse outcomes than those with T-lineage ALL (P ⫽ 0.005; data not shown). The worse treatment outcome of this subset of B-lineage patients can be attributed to the presence of the Ph chromosome (Fig. 5): Ph⫹ B-lineage patients had significantly worse outcome than that of either T-lineage or P⫺ B-lineage patients (P ⬍ 0.001 and P ⫽ 0.001, respectively). Outcome for the five patients harboring two copies of the Ph chromosome (patients 4, 5, 7, 14, and 26) did not appear to be different from that of the other Ph⫹ patients. Two of the five patients died 1.5 years or less from study entry. Three of the five patients remain alive: One underwent BMT in first remission and has remained event free 2.6 years post-BMT, one underwent BMT after a late marrow relapse and has survived 0.7 years post-BMT, and one received chemotherapy only and had survived 3.4 years. Among the three patients with monosomy 7, two had a marrow relapse, and the third died in remission due to a grampositive micrococcal infection. The patient with the high hyperdiploid clone (patient 5) also underwent BMT in first remission and is one of the event free survivors. 2036 CANCER November 1, 1998 / Volume 83 / Number 9 Event free survival for children with Ph⫹ and Ph⫺ acute lymphoblastic leukemia with white blood count (WBC) ⱖ 50,000/L at diagnosis. Number of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years were 22, 10, 9, 6, 2, and 0 for Ph⫹ and 346, 298, 261, 199, 135, and 86 for Ph⫺. FIGURE 3. FIGURE 2. Event free survival for children with Ph⫹ and Ph⫺ acute lymphoblastic leukemia according to age. (A) Age ⬍ 10 years. Number of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years were 16, 11, 10, 8, 2, and 0 for Ph⫹ and 975, 907, 837, 474, 537, and 323 for Ph⫺. (B) Age ⱖ 10 years. Number of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years were 14, 5, 5, 3, 1 and 1 for Ph⫹ and 317, 279, 254, 176, 128, and 70 for Ph⫺. Outcome After BMT for Phⴙ ALL Fifteen (50%) Ph⫹ patients in the overall cohort underwent BMT (matched sibling donor, n ⫽ 5; matched or partially matched unrelated donor, n ⫽ 7; unspecified, n ⫽ 3). Ten of the 15 patients (1882; n ⫽ 8; 1891, n ⫽ 1901, n ⫽ 1) underwent BMT in first remission that ranged from 87 days to 488 days (median 135 days). These patients ranged in age from 1 year to 16.3 years (median, 7.6 years) and had leukocyte counts ranging from 9,900/L to 800,000/L (median, 99,400/ L); all but two patients had leukocyte counts ⱖ50,000/L. Among these ten patients, four had events. Notably, the six patients (1882, n ⫽ 4; 1891, n ⫽ 1; 1901, n ⫽ 1) who remain event-free after a first remission BMT comprise the majority (6 of 8) of those Ph⫹ patients who did not experience an event; moreover, median follow-up for these six patients is 38 months (range, 24 – 60 months), suggesting that long term survival may be achieved. All of these patients, as mentioned above, remain event free after an additional 100 – 400 days of follow-up. Five patients underwent BMT after an initial relapse; three of these have died, and two remain in second remission for 0.7 and 3 years post-BMT. DISCUSSION We have determined the incidence and treatment outcome for Ph⫹ ALL in large cohort of children enrolled on contemporary, intensive risk adjusted treatment protocols of the CCG. Similar to reports by others,3–5 Ph Chromosome Positive Childhood ALL/Uckun et al. 2037 acute lymphoblastic leukemia according to early marrow status. Numbers of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years, respectively, were 12, 7, 6, 6, 2, and 1 for patients with M1 or M2 marrow status at Day 7 of induction therapy and 11, 5, 5, 2, 1, and 0 for patients with M3 marrow status at Day 7 of induction therapy. FIGURE 5. Event free survival for children ⱖ10 years of age according to immunophenotype and Philadelphia chromosome status. Percentages of 39 T lineage (T), 9 Ph⫹ B-lineage, and 25 Ph⫺ B lineage (B) patients surviving disease free during follow-up. Numbers of patients remaining in follow-up at 0, 1, 2, 3, 4, or 5 years, respectively, were 39, 34, 33, 26, 21, and 15 for T lineage; 9, 1, 1, 0, 0, 0, and 0 Ph⫹ B lineage; and 25, 20, 17, 12, 7, and 4 for Ph⫺ B lineage. Ph⫹ ALL patients comprised 2.3% of all patients with accepted to cytogenetic data. Other groups have reported that molecular genetic techniques may reveal additional cases with BCR-ABL fusion genes that were not apparent with cytogenetic analysis.21 We did not employ such techniques for diagnosis of Ph⫹ ALL in this series but are doing so for identification of Ph⫹ ALL in our ongoing trials. Any differences in outcome of patients diagnosed with molecular techniques await completion of the ongoing trials. In the cohort analyzed herein, patients were older and had higher leukocyte counts than their Ph⫺ counterparts. Unlike previous studies,3–5 we did not find an increased frequency of FAB L2 morphology but did note that Ph⫹ patients were more likely to be black than Ph⫺ patients. Consistent with previous reports,4,5 Ph⫹ ALL occurred preferentially among patients with a B-lineage immunophenotype. Cytogenetically, the majority of Ph⫹ patients were pseudodiploid, and only one patient exhibited the favorable characteristic of hyperdiploidy ⬎50. Similar to other reports,4,5,22 one patient in the current analysis had a variant Ph chromosome, and 29 patients had a balanced t(9;22)(q34;q11) either alone or in combination with other cytogenetic abnormalities. Five pa- tients harbored a second copy of the Ph chromosome, i.e., ⫹der(22)t(9;22). Although such patients with a second Ph chromosome might be expected to exhibit more aggressive disease progression than those with only a single Ph chromosome, we found no evidence for such a result in the current analysis. Monosomy 7, which has been reported to occur frequently and to confer increased risk in Ph⫹ ALL,23 was observed in only three patients, all of whom experienced events. Larger numbers of patients must be studied to determine unequivocally the significance of this cytogenetic abnormality in the context of Ph chromosome positivity. We observed a del(9p), which has been observed frequently in lymphomatous ALL24 and T-lineage ALL,25 and was associated with poor outcome in B-lineage ALL,26 in 12 of our 30 patients. It is noteworthy that only one of the eight patients without events had a deletion of 9p, whereas 11 of the 22 patients with events had a deletion of 9p. The majority of Ph⫹ patients in the current study achieved remission following induction therapy. EFS outcome, however, was poor for the overall group of Ph⫹ patients, with an estimated 4-year EFS of only 20%. Although t(9;22) was observed primarily among patients with older age and higher leukocyte counts at FIGURE 4. Disease free survival for children with Ph ⫹ 2038 CANCER November 1, 1998 / Volume 83 / Number 9 diagnosis, this translocation nevertheless was a significant risk factor for patients presenting with leukocyte counts ⬍50,000/L or age 1–9 years. Ribeiro et al.9 have reported that Ph⫹ ALL pediatric patients with leukocyte counts ⱕ25,000/L had improved outcomes compared with other Ph⫹ children. In our cohort, even Ph⫹ patients with leukocyte counts ⱕ25,000/L had a high rate of relapse. Larger numbers of patients would be required to determine the true prognostic significance of leukocyte counts in Ph⫹ ALL. In contrast to a recent report on children with Ph⫹ ALL who achieved a good response to initial prednisone therapy.10 Ph⫹ patients in the current study who had a rapid early response to induction therapy nevertheless had poor DFS outcome. The poor outcome for those with rapid early response in this series compared with other studies may be attributable to differences in treatment. For example, Arico et al. employed high dose methotrexate as part of central nervous system therapy for high risk patients including those with Ph⫹ ALL.10 Presence of the Ph chromosome appeared to be the primary factor responsible for the worse outcome observed in the CCG among older B-lineage ALL patients with high leukocyte counts compared with their T-lineage ALL counterparts. The majority of Ph⫹ ALL patients were treated on intensive chemotherapy protocols designed for higher risk patients, although treatment regimens differed. Regardless of the individual treatment regimen, outcome was uniformly poor for all Ph⫹ patients. It is noteworthy that these intensive chemotherapy programs have proven successful for the majority of patients formerly deemed high risk, including those with a T-lineage immunophenotype,27 those with lymphomatous presentation,28 those showing slow early response (SER) to induction therapy,29 and the majority of those with a t(1;19) translocation.30 Furthermore, the negative impact of the Ph chromosome apparently outweighed the previously observed, favorable effect of the CD10⫹ CD19⫹ CD34⫹ B-progenitor immunophenotype,31 because this profile was observed in the majority of our Ph⫹ patients. These findings confirm and extend reports by others that Ph⫹ ALL continues to confer adverse risk despite contemporary intensive chemotherapy regimens that yield successful outcomes for the majority of ALL patients.3,4 BMT has been employed with some success among small numbers of pediatric Ph⫹ ALL patients.3–5,32 In the current cohort, ten patients received a BMT in first remission, and it is notable that although the pretransplant conditioning regimens were not uniform, six of these ten patients remained event free for extended periods following BMT. Moreover, these six patients represent the majority (6 of 8) of those patients who have not experienced an event. These data are suggestive of a beneficial effect of first remission BMT compared with chemotherapy alone for patients with Ph⫹ ALL. In summary, the presence of a Philadelphia chromosome remains a significant adverse risk factor for childhood ALL that has not been abrogated by contemporary intensive therapies. These results emphasize the need for improved treatment programs for Ph⫹ ALL and strongly support the notion that this patient population should be considered for BMT in first remission. The application of alternative treatment strategies, such as reinforced early therapy/rotational continuation therapy,11 targeted inhibition of BCR-ABL tyrosine kinase activity,33–36 or experimental therapies, such as use of antisense BCR-ABL oligonucleotides,37 perhaps in the context of BMT, also warrant future consideration for treatment of patients with Ph⫹ ALL. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. Nowell P, Hungerford D. Chromosome studies on normal and leukemic human leukocytes. J Natl Cancer Inst 1960;25: 85. Rowley J. Ph-positive leukaemia, including chronic myelogenous leukaemia. Clin Haematol 1980;9:55. Fletcher JA, Lynch EA, Kimball VM, Donnelly M, Tantravahi R, Sallan SE. Translocation (9;22) is associated with extremely poor prognosis in intensively treated children with acute lymphoblastic leukemia. Blood 1991;77(3):435–9. Crist W, Carroll A, Shuster J, Jackson J, Head D, Borowitz M, et al. Philadelphia chromosome positive childhood acute lymphoblastic leukemia: clinical and cytogenetic characteristics and treatment outcome. a Pediatric Oncology Group study. Blood 1990;76(3):489 –94. Ribeiro RC, Abromowitch M, Raimondi SC, Murphy SB, Behm F, Williams DL. Clinical and biologic hallmarks of the Philadelphia chromosome in childhood acute lymphoblastic leukemia. Blood 1987;70(4):948 –53. Hermans A, Heisterkamp N, von Linden M, van Baal S, Meijer D, van der Plas D, et al. Unique fusion of bcr and c-abl genes in Philadelphia chromosome positive acute lymphoblastic leukemia. Cell 1987;51(1):33– 40. Clark SS, McLaughlin J, Crist WM, Champlin R, Witte ON. Unique forms of the abl tyrosine kinase distinguish Ph1positive CML from Ph1-positive ALL. Science 1987; 235(4784):85– 8. Hooberman AL, Carrino JJ, Leibowitz D, et al. Unexpected heterogeneity of Bcr-Abl fusion mRNA detected by polymerase chain reaction in Philadelphia chromosome-positive acute lymphoblastic leukemia. Proc Natl Acad Sci USA 1989; 86(11):4259 – 63. Ribeiro R, Broniscer A, Rivera G, et al. Philadelphia chromosome-positive acute lymphoblastic leukemia in children: durable responses to chemotherapy associated with low initial white blood cell counts. Leukemia 1997;11(9):1493– 6. Ph Chromosome Positive Childhood ALL/Uckun et al. 10. Arico M, Schrappe M, Harbott J, et al. Prednisone good response (PGR) identifies a subset of t(9;22) childhood acute lymphoblastic leukemia (AL) at lower risk for early leukemia relapse. Blood 1997;10:560a. 11. Roberts W, Rivera G, Raimondi S, et al. Intensive chemotherapy for Philadelphia-chromosome-positive acute lymphoblastic leukemia. Lancet 1994;343:331–2. 12. Smith M, Arthur D, Camitta B, et al. Uniform approach to risk classification and treatment assignment for children with acute lymphoblastic leukemia [see comments]. J Clin Oncol 1996;14(1):18 –24. 13. Steinherz PG, Siegel SE, Bleyer WA, et al. Lymphomatous presentation of childhood acute lymphoblastic leukemia. Cancer 1991;68:751– 8. 14. Heerema NA, Palmer CG, Baehner RL. Karyotypic and clinical findings in a consecutive series of children with acute lymphocytic leukemia. Cancer Genet Cytogenet 1985;17(2): 165–79. 15. Mitelman F. ISCN: an international system for human cytogenetic nomenclature. Basel, Switzwerland: Karger, 1995. 16. Uckun FM, Muraguchi A, Ledbetter JA, et al. Biphenotypic leukemic lymphocyte precursors in CD2⫹CD19⫹ acute lymphoblastic leukemia and their putative normal counterparts in human fetal hematopoietic tissues. Blood 1989; 73(4):1000 –15. 17. Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457– 81. 18. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163–70. 19. 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 1977; 35:1. 20. Nachman J, Sather H, Cherlow J, Sensel MG, Gaynon PS, Lukens JN et al. Response of children with high-risk acute lymphoblastic leukemia treated with and without cranial irradiation: a report from the Children’s Cancer Group. J Clin Oncol 1998;16:920 –30. 21. Westbrook C, Hooberman A, Spino C, et al. Clinical significance of the BCR-ABL fusion gene in adult acute lymphoblastic leukemia: a cancer and leukemia Group B study (8762). Blood 1992;80:2983–90. 22. Dow LW, Tachibana N, Raimondi SC, Lauer SJ, Witte ON, Clark SS. Comparative biochemical and cytogenetic studies of childhood acute lymphoblastic leukemia with the Philadelphia chromosome and other 22q 11 variants. Blood 1989; 73(5):1291–7. 23. Russo C, Carroll A, Kohler S, et al. Philadelphia chromosome and monosomy 7 in childhood acute lymphoblastic leukemia: a Pediatric Oncology Group Study. Blood 1991;77(5): 1050 – 6. 24. Chilcote RR, Brown E, Rowley JD. Lymphoblastic leukemia with lymphomatous features associated with abnormalities of the short arm of chromosome 9. N Eng J Med 1985;313: 286 –91. 2039 25. Raimondi SC, Behm FG, Roberson PK, et al. Cytogenetics of childhood T-cell leukemia. Blood 1988;72(5):1560 – 6. 26. Rivera GK, Raimondi SC, Hancock ML, et al. Improved outcome in childhood acute lymphoblastic leukaemia with reinforced early treatment and rotational combination chemotherapy. Lancet 1991;337(8733):61– 6. 27. Uckun F, Reaman G, Steinherz P, et al. Improved outcome for children with T-lineage acute lymphoblastic leukemia after contemporary chemotherapy: a Children’s Cancer Group study. Leuk Lymphoma 1996;24:57–70. 28. Altman A, Steinherz P, Trigg M, et al. A prospective randomized trial of granulocyte colony-stimulating factor (G-CSF) during induction or consolidation phases of intensive chemotherapy for acute lymphocytic leukemia (ALL) [meeting abstract]. J Clin Oncol 1996;15. 29. Nachman J, Sather HN, Gaynon PS, Lukens JN, Wolff L, Trigg ME. Augmented Berlin-Frankfurt-Munster therapy abrogates the adverse prognostic significance of slow early response to induction chemotherapy for children and adolescents with acute lymphoblastic leukemia and unfavorable presenting features: a report from the Children’s Cancer Group. J Clin Onol 1997;15:2222–30. 30. Uckun F, Sensel M, Sather H, et al. Clinical significance of translocation t(1;19) in childhood acute lymphoblastic leukemia in the the context of contemporary therapies: a report from the Children’s Cancer Group. J Clin Oncol 1997;16:527– 35. 31. Uckun F, Sather H, Gaynon P, et al. Prognostic significance of the CD10⫹CD19⫹CD34⫹ immature B-progenitor immunophenotype in children with acute lymphoblastic leukemia: a report from the Children’s Cancer Group. Leuk Lymphoma 1997;27:445–57. 32. Sierra J, Radich J, Hansen JA, et al. Marrow transplants from unrelated donors for treatment of Philadelphia chromosome-positive acute lymphoblastic leukemia. Blood 1997; 90(4):1410 – 4. 33. Carlo Stella C, Dotti G, Mangoni L, et al. Selection of myeloid progenitors lacking BCR/ABL mRNA in chronic myelogenous leukemia patients after in vitro treatment with the tyrosine kinase inhibitor genistein. Blood 1996;88(8):3091– 100. 34. Druker BJ, Tamura S, Buchdunger E, et al. Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 1996;2(5):561– 6. 35. Honma Y, Matsuo Y, Hayashi Y, Omura S. Treatment of Philadelphia-chromosome-positive human leukemia in SCID mouse model with herbimycin A, bcr-abl tyrosine kinase activity inhibitor. Int J Cancer 1995;60(5):685– 8. 36. Uckun FM, Evans WE, Forsyth CJ, et al. Biotherapy of B-cell precursor leukemia by targeting genistein to CD19-associated tyrosine kinases. Science 1995;267(5199):886 –91. 37. Skorski T, Nieborowska Skorska M, Wlodarski P, et al. Treatment of Philadelphia leukemia in severe combined immunodeficient mice by combination of cyclophosphamide and bcr/abl antisense oligodeoxynucleotides [see comments]. J Natl Cancer Inst 1997;89(2):124 –33.