Leukemia (2001) 15, 1038–1045  2001 Nature Publishing Group All rights reserved 0887-6924/01 $15.00 www.nature.com/leu A comparison of early intensive / with early intensive alternating combination for high-risk B-precursor acute lymphoblastic leukemia: a Pediatric Oncology Group phase III randomized trial SJ Lauer1, JJ Shuster2, DH Mahoney Jr3, N Winick4, S Toledano5, L Munoz5, G Kiefer6, JD Pullen7, CP Steuber3 and BM Camitta6

1Emory University School of Medicine, Atlanta, GA; 2Pediatric Oncology Group Statistical Office and Department of Statistics, University of Florida, Gainesville, FL; 3Texas Children’s Center, Baylor College of Medicine, Houston, TX; 4University of Texas Southwestern Medical Center, Dallas, TX; 5University Miami School of Medicine, Miami, FL; 6Midwest Children’s Cancer Center, Milwaukee, WI; and 7University of Mississippi Medical Center Children’s Hospital, Jackson, MS, USA

A prospective, randomized multicenter study was performed to to their risk for relapse and treatment strategies designed to evaluate the relative efficacy of two different concepts for early improve event-free survival (EFS). intensive therapy in a randomized trial of children with B-pre- cursor acute lymphoblastic leukemia (ALL) at high risk (HR) for It is believed that the leading causes of relapse in children relapse. Four hundred and ninety eligible children with HR-ALL with higher risk ALL (HR-ALL) are inadequate cell kill and were randomized on the Pediatric Oncology Group (POG) 9006 emergence of drug-resistant clones. Clinical trials using early phase III trial between 7 January 1991 and 12 January 1994. intensive myelosuppressive combination chemotherapy as After prednisone (PDN), (VCR), (ASP) post-induction consolidation were designed to maximize cell and daunorubicin (DNR) induction, 470 patients received either 2 kill and address drug resistance. Subsequently, improvement 12 intensive parenteral treatments of intermediate dose (1 g/m 9–20 each) methotrexate (MTX) and mercaptopurine (MP) over 24 in survival was realized for patients with HR-ALL. weeks (regimen A) or 12 intensive course of alternating myelos- Increased toxicity was also noted with the aggressive use of uppressive drug combinations given over 30 weeks (regimen myelosuppressive agents.9,13,14,16,18,19–22 However, prelimi- B). These drug combinations included MTX/MP, nary results from POG 8698 suggested that early intensifi- (VM-26)/cytosine arabinoside (AC) and VCR/PDN/DNR/AC/ASP. cation with the less toxic combination of intermediate-dose Central nervous system (CNS) prophylaxis was age-adjusted methotrexate (MTX) and mercaptopurine (MP) might be as triple intrathecal chemotherapy. Patients with CNS disease at effective as the more myelosuppressive combinations used in diagnosis were treated with craniospinal irradiation after the 13,17 intensive phase. Continuation was standard doses of MTX and another POG pilot study (POG 8398) for HR-ALL. MP for 2 years. This trial was closed early because of an appar- In 1991, the POG opened a group-wide randomized phase ent early difference favoring regimen B. Results show that 470 III (POG 9006) to treat children with HR-ALL. The patients achieved remission (97%). Two hundred and thirty two primary objectives of this randomized trial were to compare were randomized to regimen A and 238 to regimen B. The esti- the efficacy and toxicity of regimen A: 12 early intensive mated 4-year event-free survival (EFS) for patients treated with regimen A is 61.6 % (s.e. = 3.3%) and with regimen B is 69.4% courses of -based chemotherapy (intermediate- (s.e. = 3.1%), P = 0.091. Toxicities were more frequent on regi- dose MTX/MP) vs regimen B: 12 early intensive courses of men B. In conclusion, for children with B-precursor ALL at high alternating myelosuppressive, non-cross-resistant combination risk to relapse, early intensification with myelosuppressive chemotherapy with MTX/MP as per the Goldie–Coldman combination chemotherapy was more toxic but produced no hypothesis.23 This paper reports the results of that trial. significant difference in EFS when compared to those treated with parenteral methotrexate and mercaptopurine. Leukemia (2001) 15, 1038–1045. Keywords: high risk childhood ALL; B-precursor ALL; early inten- Patients and methods sive chemotherapy Patients

The POG 9006 phase III clinical trial accrued patients between 7 January 1991 and 12 January 1994. Approval by Introduction local institutional review boards and written informed consent were required before patient entry. Improvement in overall survival of children with acute lym- phoblastic leukemia (ALL) is in part due to a better under- standing of the biology of this malignancy and factors that are Eligibility associated with the success or failure to maintain remission. Irrespective of the therapeutic trial, common risk factors Eligibility for POG 9006 included (1) enrollment on the POG include age, WBC at diagnosis, leukemic cell surface antigen 9000 classification study; (2) confirmation of B-precursor ALL 24 expression, cytogenetics, DNA index and early response to by central reference laboratories; and (3) meeting the criteria cytoreductive therapy.1–8 Based on these prognostic factors, for high risk B-precursor ALL. Those criteria were leukemic < children with newly diagnosed ALL can be stratified according cell DNA index of 1.16 (DNA content in leukemic cells: DNA content of normal G0/G1 cells) (DI) by central reference laboratory25,26 and at least one of the following: (1) WBC >10 000–99 000/µl, aged 1–2.99 years or ages 6–21 years; Correspondence: SJ Lauer, c/o Pediatric Hematology-Oncology, 2040 > µ Ridgewood Drive, Suite 100, Atlanta, GA 30322, USA; Fax: 1 404 (2) WBC 100 000/ l, aged 1–21 years; (3) all patients with 727 4455 CNS or overt testicular disease at diagnosis; or (4) leukemic Received 16 November 2000; accepted 26 January 2001 cell chromosome translocations t(1;19) or t(9;22) confirmed Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1039 by central reference laboratory.27,28 Only patients who met Treatment and drug dose modification criteria 1, 2, or 3 were eligible for randomization. Patients having t(1;19), or t(9;22) leukemia at diagnosis were not ran- Patients were randomized at diagnosis to one of two intensi- domized because the numbers of these patients were pre- fication schedules which have been previously outlined in dicted to be low and thus no statistically valid information detail.13,17 Treatment regimens are listed in Table 2. Induction would be obtained if they were randomized. All of these therapy was identical for both groups: vincristine (VCR), patients were assigned to regimen A and were excluded from prednisone (PDN), E. coli l-asparaginase (ASP) and daunorub- this report. Patients ,12 months of age were not eligible for icin (DNR). Age-adjusted triple intrathecal therapy (TIT) with this protocol. methotrexate (MTX), hydrocortisone (HDC) and cytosine arab- inoside (AC) was administered on day 1 of induction. Patients with CNS disease at diagnosis were given three additional Definition of disease and response weekly doses of age-adjusted intrathecal MTX. Intensification started immediately after meeting the criteria for remission. CNS leukemia was diagnosed when the cerebrospinal fluid Patients randomized or assigned regimen A began week 1 of (CSF) WBC count was >5 cells/µl and lymphoblasts were intensification receiving intravenous (i.v.) intermediate-dose identified on a Wright-stained, cytocentrifuged slide examin- MTX infused over 24 h, followed by i.v. intermediate-dose MP ation without peripheral blood (PB) contamination. Complete infused over 6 h. On week 2, patients received intramuscular remission (CR) was defined as a cellular bone marrow with (i.m.) MTX on day 1 and MP by mouth (p.o.) daily for 7 days. fewer than 5% blasts and no evidence of leukemia at any The 2-week schedule was repeated 12 times over 24 weeks. other site. The i.v. administration of MTX and MP required a 48-h hospi- The definition of relapse required: (1) bone marrow .25% talization. Those patients randomized to regimen B received lymphoblasts; (2) CNS >5 WBC/µl of CSF and lymphoblasts six courses of MTX and MP as in regimen A, three courses of identified on a Wright-stained examination without PB con- teniposide (VM-26) and AC and three courses of DNR, AC, tamination; (3) extramedullary site with biopsy proven infil- VCR, PDN and PEG-asparaginase (PEG-ASP) in an alternating trate with lymphoblasts; or (4) any combination of the above. fashion over 30 weeks. The DNR/AC and VM-26/AC courses required a 72-h hospitalization. Plasma MTX levels were monitored after each intermediate-dose MTX. Leucovorin Randomization (LCV) rescue began 48 h after the start of the methotrexate infusion and continued every 6 h for five doses or until plasma Five hundred and seventy-three newly diagnosed patients with MTX was ,0.1 µmol/l. All courses of chemotherapy during B-precursor, HR-ALL were registered on POG 9006. After intensification began when the absolute neutrophil count was eligibility for this HR-ALL protocol was determined, 69 >500/µl and platelet count was >100 000/µl. If courses of patients (45 t(1;19); 24 t(9:22)) were removed from randomiz- DNR/AC or VM-26/AC resulted in prolonged neutropenia ation and assigned regimen A, 14 patients were ineligible for (,500/µl for .24 days), each myelosuppressive drug was protocol therapy (wrong diagnosis) and 490 patients were reduced by 25% for the next course. Dose escalation was not randomized (Table 1) to receive one of two post-induction permitted during any phase of this protocol. intensification therapies (Table 2). Because of the known sensitivity of Down syndrome patients to MTX and myelosuppressive chemotherapy, they Table 1 Presenting patient characteristics started with a 50% dose reduction of intermediate-dose MTX, Characteristics Regimen A Regimen B DNR, AC and VM-26 during intensification. Subsequent n = 243 n = 247 courses of these agents were increased or decreased accord- ing to tolerance. Female:male 106:137 106:141 CNS prophylaxis was continued throughout intensification Age (years)a 7.9 (2.5–13.3) 7.9 (2.5–13.4) and continuation with age-adjusted TIT for a total of 18 doses Caucasian 183 181 for both regimens. Those children with CNS disease at diag- Black 18 22 nosis received additional IT chemotherapy (MTX) during Hispanic 30 32 induction (as above) and eight doses of TIT during intensifi- Other 12 12 WBCa 25 (11–83) 22 (8–62) cation (regimen A or B). Following intensification these CNS disease + other EMDb 710patients then received craniospinal irradiation: cranial vol- NCI good vs poor 85:158 79:168 ume, 2400 cGy in 16 fractions; and spinal axis, 1500 cGy in Subgroups 10 fractions. No intrathecal chemotherapy was given follow- Down syndrome 3 5 ing irradiation. t(4;11) 8 6 Following intensification all patients were given identical t(1;19)c 45 NR t(9;22)c 24 NR continuation therapy with standard-dose MTX and MP. Inten- sification and continuation lasted for a total of 130 weeks. n, number; WBC, white blood count per µl; CNS, central nervous system; EMD, extramedullary disease; NCI, National Cancer Insti- tute consensus risk group definition; NR, not randomized. Statistical considerations aMedian (quartiles) b15 patients with CNS disease at diagnosis, one patient with tes- The plan for this study was to randomize 507 patients and ticular disease at diagnosis, one patient with eye disease at diag- monitor the patients until the last entrant would be at risk for nosis; three patients with CNS disease at diagnosis had t(1;19) and one patient had t(9;22). 4 years. This plan allowed greater than 90% power to detect ct(1;19), t(9;22) patients were not part of the therapeutic question a 12% difference in 4-year continuous complete remission for the clinical trial and are not included in the 243 patients on (CCR) rates (60% vs 72%), based on a two-sided logrank test regimen A. at P = 0.05, proportional hazards, and a post-4-year hazard

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1040 Table 2 Treatment regimens

Induction (all patients)

Vincristine 1.5 mg/m2 (max 2 mg) i.v. weekly ×4 Prednisone 40 mg/m2 (max 60 mg) p.o. daily ×28 divided three times a day Asparaginase 6000 IU/m2 i.m. days 2, 5, 8, 12, 15, 19 Daunorubicin 30 mg/m2 i.v. days 8, 15, 22 TIT day 1 (age-adjusted), CNS disease IT-MTX (age-adjusted) days 8, 15, 22

Intensification (randomization)

Regimen A (wk 1–24) Regimen B (wk 1–31) Week 1 Wk 1, 6, 11, 16, 21, 26 MTX 1000 mg/m2 i.v. over 24 h, then Same as Regimen A MP 1000 mg/m2 i.v. over 6 h Wk 3, 13, 23 LCV 5 mg/m2 i.v. or p.o. every 6 h × 5a VM-26 165 mg/m2 i.v. days 1, 2 Week 2 AC 150 mg/m2/24 h i.v. or s.c CI × 3 days MTX 20 mg/m2 i.m. day 1 Wk 8, 18, 28 MP 50 mg/m2 p.o. days 1–7 DNR 30 mg/m2 i.v. days 1, 14 Repeat 2-wk cycles for total of 12 courses VCR 1.5 mg/m2 i.v. days 1, 8 PDN 40 mg/m2 p.o. day 1–7 PEG-ASP 2500 IU/m2 i.m. day 1 AC 150 mg/m2/24 h i.v. on s.c. CI × 3 days

Continuation wk 25–130 (all patients)

MTX 20 mg/m2 i.m. day 1 MP 50 mg/m2 p.o. days 1–7

CNS Therapy: TITb (age adjusted) Intensification Regimen A Regimen B Wk 1, 2, 3, 7, 11, 15, 21, 25, 31 Wk 1, 2, 3, 6, 11, 16, 21, 26, 31 CNS disease: Craniospinal XRT, 2400 cGy/1500 cGy

Continuation: (all patients except CNS disease at diagnosis) Wk 42, 55, 67, 79, 91, 104, 116, 128

Age (years) 1 2 >3 >9

MTX (mg) 8 10 12 15 HDC (mg) 8 10 12 15 Ac (mg) 16 20 24 30

p.o., by mouth; max, maximum; i.v. intravenous; i.m., intramuscular; TIT, triple intrathecal therapy; CNS, central nervous system; wk, week; s.c., subcutaneous; CI, continuous infusion; MTX, methotrexate; MP, mercaptopurine; LCV, leucovorin, VM-26, teniposide; AC, cytosine arabinoside; DNR, daunorubicin; VCR, vincristine; PDN, prednisone, PEG-ASP, PEG-asparaginase; XRT, X-ray therapy; HDC, hydrocorti- sone. aStarts 48 h from start of MTX and continues for a minimum of five doses or until serum MTX ,0.1 µmol/l. bHalf-dose i.m.-MTX when given on same days as TIT.

of 25% of the pre-4 year hazard.29 The Data Monitoring Com- Readers are cautioned against overinterpretation of subsets mittee (January 1994) closed the trial for an apparent early and site-specific failure comparisons. The overall results difference favoring regimen B (CCR at 2 years, A vs B, 70.8% should take priority in all subsets, because the study was not (s.e. 7.7%) vs 82% (s.e. 6.1%), P = 0.0016. Accrual at the time planned for these secondary analyses from a statistical power of study closure was 490, 17 less than the planned accrual of perspective. No subset demonstrated a qualitative interaction, 507). The committee recommended that the data be allowed where a result in favor of the overall inferior treatment to mature to the planned follow-up before publication. occurred. We performed a Cox analysis32 to test for a quanti- Since both regimens used the same induction therapy, the tative interaction between treatment and sex. This tests for the primary end point was CCR, the time from achievement of a equality of the treatment effect size between males and CR to failure (, relapse, or second malignancy) or last females. contact. Event-free survival (EFS) results and site-specific fail- ure results are also presented (Table 3). EFS is similar to CCR, except that the clock starts at registration and induction fail- Results ures are counted. Actuarial comparisons were conducted by the logrank test. Actuarial curves were constructed by the Induction method of Kaplan–Meier30 using standard errors of Peto et al.31 The cutoff for analysis was October 1998, the earliest Four hundred and ninety (490) patients were randomized to cutoff where the planned follow-up was completed in all receive one of two post-induction therapies. Twenty patients patients. did not achieve complete remission (CR) status: 15 due to

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1041 Table 3 Outcome by treatment

Variable Regimen A Regimen B P valuea

n Fail Expected n Fail Expected

EFS 243 98 86.5 247 86 97.5 0.091 CCRa 232 88 80.0 238 81 89.0 0.22 Marrow 232 64 56.2 238 55 62.8 0.15 CNS 232 25 26.7 238 31 29.3 0.64 Testicular 130 11 8.3 137 8 10.7 0.22 Male (CCR) 130 62 51.0 137 50 61.0 0.036 Female (CCR) 102 26 28.4 101 31 28.6 0.52 Caucasian (CCR) 175 65 60.9 174 61 65.1 0.46 Black (CCR) 18 5 5.1 21 6 5.9 0.95 Hispanic (CCR) 28 12 10.6 31 11 12.4 0.55 Other (CCR) 11 6 — 12 3 — — n, number; EFS, event-free survival; CCR, continuous complete remission; CNS, central nervous system. aThe major end point is CCR, since the two arms had the same induction regimen. All P values are by two-sided logrank analysis. induction failure and five were not evaluable for CR (refused estimated 4-year CCR rate for overall CNS (isolated and therapy, one; toxicity, one; non-documentation, three) for a combined) relapse is 87.7% (s.e. = 3.3%) for regimen A and remission induction rate of 97% (470/485). Thus 470 of 485 85.4% (s.e. = 3.1%) for regimen B, P = 0.64. The estimated patients achieved complete remission and were eligible for 4-year CCR rate for marrow relapse is 70.9% (s.e. = 4.1%) for the randomized study question (regimen A, 232/243; regimen regimen A and 75.3% (s.e. = 3.6%) for regimen B, P = 0.15. B, 238/247). No significant difference in the incidence of testicular relapse by treatment regimen was observed, P = 0.22. The estimated overall EFS for the 16 randomized patients with CNS disease Outcome at diagnosis is 46.2% (s.e. = 16.9%). Only one of these patients had CNS involvement at relapse (combined BM + Patient outcomes by regimen are shown in Figures 1 and 2. CNS). There was a higher failure rate for males than for The 4-year estimated EFS rate for randomized eligible patients females treated on regimen A. Although the treatment differ- treated with regimen A is 61.6% (s.e. = 3.3%) and with regi- ence was significant (P = 0.036) for males and not statistically men B is 69.4% (s.e. = 3.1%), P = 0.091. The 4-year estimated significant for females (P = 0.52), one should bear in mind overall CCR rate for patients with regimen A is 64% (s.e. = that this was not a predesigned question. Also Cox analysis 3.4%) and with regimen B is 70.6% (s.e. = 3.1%), P = 0.22. was conducted to compare the treatment effect size (hazards This study was inconclusive with respect to efficacy. Based ratio) within males vs that within females. The estimated ratio on the 5-year CCR, we are 95% confident that the true differ- of hazards ratios is 176% (95% confidence limits, 92% to ence ranges from 6% favoring regimen A to 15% favoring regi- 469%) males:females. Equivalent treatment effects (100%) fall men B. Logrank comparisons of treatment outcome by sites of within the confidence interval. Since no interaction could be failure and within gender and racial subgroups are listed in demonstrated by this Cox analysis, the overall results should Table 3 and accounting of events is listed in Table 4. The take priority over sex-specific results.

Figure 1 Kaplan–Meier plot of the probability of event-free survival (EFS) for patients randomized to regimen A or B. P, estimated percent of patients failure-free to the end of interval; SE, standard error of P; F, number of failures in interval; N, number at risk at start of interval.

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1042

Figure 2 Kaplan–Meier plot of the probability of complete continuous remission (CCR) for patients randomized to regimen A or B. P, estimated percent of patients failure-free to the end of interval; SE, standard error of P; F, number of failures in interval; N, number at risk at start of interval.

Table 4 Accounting of events patients) during the intensification phase. Regimen B had a 26% higher incidence of severe neutropenia when compared Regimen A Regimen B to regimen A. Hospitalizations for fever and neutropenia were 25% more frequent for patients treated with regimen B vs regi- Patients 243 247 men A. Documented bacterial sepsis was 7% more frequent Males with CR 130 137 in regimen B. Drug fevers were also more frequent in regimen Not CR 10 5 Marrow 64a 55b B due to the 72 h AC infusions. Allergic drug reactions to VM- CNS 25a 31b 26 and ASP were isolated to regimen B. There were five Isolated CNS 16 18 during remission: two in regimen A (cardiac, one; Testicular 11a 8b infection, one) and three in regimen B (liver failure, one; infec- Isolated testicular 5 3 tion, two). Therapy was discontinued permanently during Unknown site 0 1 treatment due to toxicity for 11 patients on regimen A and 22 Remission deaths 2 3 = Second 1c 0 patients on regimen B (P 0.047, 11/232 vs 22/238, exact Total failure 98 86 conditional chi-square). For regimen A, 153 patients modified or omitted a component of therapy to deal with toxicity vs aOf the 64 marrow relapses, nine had CNS involvement and six had 162 for regimen B. testicular involvement. bOf the 55 marrow relapses, 12 had CNS involvement and four had testicular involvement. One non-marrow relapse was jointly CNS Neurotoxicities and testicular. cAstrocytoma (brain). The incidences of neurotoxic events (NTE) are presented in Table 6. There were 61 patients who had one or more grades Toxicities 3–4 NTE (10.9%). The incidence was comparable between regimens A and B (31 vs 30). Seizures were the most common Common grades 3–4 toxicities are listed in Table 5. Significant event (32 of 61 neurotoxic events). Clinicians judged the neur- neutropenia was the most common toxicity recorded (70% of otoxic event to be methotrexate-associated (MTX-NTE) in 56 of the 61 patients (10% of all patients at risk). Of 31 patients Table 5 Toxicity profiles of intensification regimens with neurotoxicity who had brain MRI or CT scans following their event, 19 (61%) had imaging evidence for white matter a Type of toxicity Regimen A Regimen B changes/leukoencephalopathy. Eight patients were removed n = 232 (%) n = 238 (%) from this therapeutic trial because of unacceptable neurotoxicity. Neutropeniab 130 (56) 192 (82) Thrombocytopenia 22 (9) 59 (25) Anemia 6 (3) 20 (8) Fever/neutropenia 19 (8) 78 (33) Subgroup analyses Bacterial sepsis 20 (9) 37 (16) Stomatitis 105 (45) 117 (49) Outcome for randomized subgroups include 14 patients with Drug fever 0 (0) 12 (5) Allergic reactionc 2 (1) 52 (22) t(4;11): seven relapsed (CNS, three; BM, four), two went to bone marrow transplantation (BMT) and five remain in CR; aGrades 3–4 toxicity reported. Toxicity was recorded only once eight Down syndrome patients: two died during induction per patient. (sepsis), one lost to follow-up and five remain in CR. Details bGrades 3–4 neutropenia: 500–900 or ,500 µl. of patients with t(1;19) or t(9;22) will be reported as part of a cAllergic reactions to VM-26 and/or asparaginase. larger POG experience.

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1043 Table 6 Neurotoxicity profiles for all patients (n = 559)a vs 8, P = 0.22). Since gender was not a predesigned study question, the overall results take priority. Type of toxicity Frequencyb (%) Regimen B was clearly more toxic than regimen A (Table 5). Grades 3–4 toxicities related to cytopenic events were 5% Seizures 32 (5.7) to 26% higher in regimen B. Hospitalizations for Paresis 10 (1.8) fever/neutropenia and bacterial sepsis were 25% and 7% Vincristine PN 2 (0.4) Hemorrhage/thrombosis 1 (0.2) more prevalent for regimen B. Drug fevers and allergic reac- Otherc 16 (2.9) tions were much higher for regimen B. There were 5 deaths Totald 61 (10.9) while in remission (infectious (three), cardiac (one), and liver failure (one)), two on regimen A, and three on regimen B. Subgroup analysis The only second malignancy reported among the randomized e Regimen A 31 patients was a brain tumor in a patient with CNS disease at Regimen Bf 30 MTX-associated neurotoxic event (MTX-NTE) 56/559 (10.0) diagnosis (regimen A) and who received craniospinal Leukoencephalopathy by CT/MRIg 19/31 (61) irradiation. Acute neurotoxic events (NTE) (grades 3–4) were similar PN, peripheral neuropathy; MTX, methotrexate; CT, computer tom- between regimens (Table 6). The overall methotrexate-asso- ography; MRI, magnetic resonance imaging. ciated NTE was 10%. These results are similar to those a559 patients = 490 randomized + 45 t(1;19) + 24 t(9;22) patients. reported for the POG 9005 (standard risk ALL) where inter- bEvent is recorded only once per patient. mediate-dose MTX and TIT were used in a similar fashion.33,34 cSevere headaches, slurred speech, ataxia, parethesias. d Potential reasons for these neurotoxicities have been reviewed Total number of patients with at least one grade 3–4 neurotoxic 34 event. in a previous publication, but include the number of doses eRegimen A, 232 randomized + 45 t(1;19) + 24 t(9;22) patients = of i.v. intermediate dose MTX, the concomitant use of inter- 301. mediate-dose MTX and TIT during intensification, the ratio of fRegimen B, 238 randomized. intermediate-dose MTX to leucovorin rescue and/or the lack gFrequency of white matter changes identified on CT or MRI brain of leucovorin following TIT during continuation. Patients in scans over the total number of patients with imaging studies regimen A received 12 courses of intermediate-dose MTX in reported. 24 weeks while those in regimen B received six courses in 30 weeks, yet the incidence of methotrexate associate NTE were similar (31 vs 30). This observation could be explained by the Discussion fact that both regimens prescribed the same number of TITs which, during the intensive phase, were only given during Early intensification is designed to continue cell kill and pre- intermediate-dose MTX administration. vent the emergence of drug-resistant leukemia as a cause of Overall outcome data from this clinical trial compare favor- treatment failure. The use of alternating myelosuppressive ably with previous POG trials and those of other groups treat- combination chemotherapy early in the post-induction period ing children for HR-ALL (62–75%).9–20 Because risk group cri- was tested in the POG 8398 pilot protocol.13 Significant but teria differ among large cooperative groups treating childhood tolerable toxicity was encountered. Drug combinations were ALL, external comparisons are difficult and hazardous. How- selected for their antileukemic effects in relapsed disease and ever, 66% of the patients in this study would be considered for their relative non-cross-resistance as suggested by the Gol- high risk by the CTEP/NCI consensus risk group definition (age die–Coldman hypothesis.23 A more complete understanding >10 years, or WBC >50 000) thus adding some validity to of the mechanisms of drug resistance and cell kill prompted cross comparisons.35 The data show that sites of relapse are the addition of vincristine, prednisone and asparaginase to the also consistent with other trials where bone marrow was drug combination daunorubicin/Ara-C. The VM-26 and Ara- identified as the primary site of failure. The incidence of iso- C combination alternating with IDMTX/MP remained the lated CNS relapse was 7% (34/490) and equivalent between same as in the POG 8398 pilot study. This intensive combi- regimens. As in previous POG trials, intrathecal and systemic nation was compared in a randomized trial to the less toxic chemotherapy provide excellent CNS prophylaxis and avoid but equally efficacious anti-metabolite combination MTX/MP the use of cranial irradiation.16,34–36 as supported by the POG 8698 pilot study.17 Recent trials from BFM, CCG and MRC UKALL reported Results from this phase III trial were inconclusive with improved results for HR-ALL patients when compared to pre- respect to efficacy. That is, children with HR-ALL treated with vious trials.14,19,37 Improvement was attributed to the use of early intensive therapy using intermediate-dose MTX/MP blocks of intensive therapy (consolidation, intensification, alone showed no significant difference in EFS or CCR when VCR plus PDN pulses) over the first year of remission vs compared to those similarly treated with alternating myelos- POG’s intensive therapy limited to the first 6 months of uppressive combinations. The 4-year estimated EFS and CCR treatment. for the intermediate-dose MTX/MP alone compared to the Reasons for lack of a significant difference between the two multidrug combinations were 61.6 (3.3%) vs 69.4% (s.e. = regimens may be attributed to one or more of the following: 3.1%), P = 0.091 and 64% (s.e. = 3.4) vs 70.6 % (s.e. = 3.1%), (1) the true difference may be less than the planned differences P = 0.22, respectively. The randomized groups were matched making the study power inadequate to be sensitive to the true for the risk factors age, gender, WBC and DI (Table 1). When difference; (2) using multiple blocks of intensification may be outcome was compared between treatment regimens for sites more important in improving outcome than the specific agents of relapse and ethnicity, no significant differences were found used;21,37,38 (3) the alternating drug combinations did not (Table 3). However, for males the failure rate was lower in adequately test the Goldie–Coldman hypothesis, ie the drugs the alternating arm compared to the intermediate-dose selected were not equally effective nor non-cross-resistant. It MTX/MP arm (50 events vs 62 events) with a P = 0.036. This is now well established that DNR and VM-26 share similar difference is not related to an increase in testicular relapse (11 mechanisms of cell kill (topoisomerse II inhibition) and of

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1044 resistance (multidrug resistance).39–41 However, this clinical presenting features: A follow-up report of the Children’s Cancer trial demonstrated that comparable efficacy can be achieved Group Study CCG-106. J Clin Oncol 1993; 11: 2234–2242. with antimetabolite therapy alone while avoiding many of the 13 Lauer SJ, Camitta BM, Leventhal BG, Mahoney D, Shuster JJ, Adair S, Casper J, Civin C, Graham M, Keifer G, Pullen J, Steuber P, early and later toxicities of more myelosuppressive agents. Kamen B. Intensive alternating drug pairs for treatment of high- Better treatment is still needed for patients at higher risk of risk childhood acute lymphoblastic leukemia. Cancer 1993; 71: relapse. 2854–2861. 14 Reiter A, Schrappe M, Ludwig WD, Hiddemann W, Sauter S, Henze G, Zimmermann M, Lampert F, Havers W, Niethammer D, Acknowledgements Odenwald E, Ritter J, Mann G, Welte K, Gadner H, Riehm H. Chemotherapy in 998 unselected childhood acute lymphoblastic leukemia patients. Results and conclusions of the multicenter trial This work was supported in part by the following grants from ALL-BFM 86. Blood 1994; 84: 3122–3133. the National Cancer Institute, Bethesda, MD: CA 20549, CA 15 Schorin MA, Blattner S, Gelber RD, Tarbell NJ, Donnelly M, 29139, CA 03161, CA 33625, CA 32053, CA 15989, CA Dalton V, Cohen HJ, Sallan SE. Treatment of childhood acute 30969. lymphoblastic leukemia: Results of Dana-Farber Cancer Institute/Children’s Hospital acute lymphoblastic leukemia con- sortium Protocol 85–01. J Clin Oncol 1994; 12: 740–747. 16 Land VJ, Shuster JJ, Crist WM, Ravindranath Y, Harris MB, Krance References RA, Pinkel D, Pullen DJ. Comparison of two schedules of inter- mediate-dose methotrexate and consolidation therapy 1 Pullen DJ, Crist WM, Falletta JM, Vogler LB, Dowell B, Humphrey for childhood B-precursor cell acute lymphoblastic leukemia: A GB, Blackstock R, Eys JV, Cooper MD, Metzgar RS, Meydrech EF. Pediatric Oncology Group Study. J Clin Oncol 1994; 12: 1939– Southwest Oncology Group experience with immunological 1945. phenotyping in acute lymphocytic leukemia of childhood. Cancer 17 Camitta BM, Mahoney D, Levethal BG, Lauer SJ, Shuster JJ, Adair Res 1981; 41: 4802–4809. S, Civin C, Munoz L, Steuber P, Strother D, Kamen BA. Intensive 2 Look AT, Roberson PK, Williams DL, Rivera G, Bowman WP, Pui intravenous methotrexate and mercaptopurine treatment of high- CH, Ochs J, Abromowitch M, Kalwinsky D, Dahl GV, George S, risk non-T acute lymphocytic leukemia. J Clin Oncol 1994; 12: Murphy SB. Prognostic importance of blast cell DNA content in 1383–1389. childhood acute lymphocytic leukemia. Blood 1985; 65: 1079– 18 Winick N, Shuster JJ, Bowman WP, Borowitz M, Farrow A, Jaca- 1086. ruso D, Buchanan GR, Kamen BA. Intensive oral methotrexate 3 Miller DR, Coccia PF, Bleyer WA, Lukens JN, Siegel SE, Sather protects against lymphoid marrow relapse in childhood B-precur- HN, Hammond GD. Early response to induction therapy as a pre- sor acute lymphoblastic leukemia. J Clin Oncol 1996; 14: dictor of disease-free survival and late recurrence of childhood 2803–2811. acute lymphoblastic leukemia: A report from the Children’s Can- 19 Nachman J, Sather HN, Cherlow JM, Sensel MG, Gaynon PS, cer Study Group. J Clin Oncol 1989; 7: 1807–1815. Lukens JN, Wolff L, Trigg ME. Response of children with high-risk 4 Pui C-H, Crist WM, Look AT. Biology and clinical significance of cytogenetic abnormalities in childhood acute lymphoblastic leu- acute lymphoblastic leukemia treated with and without cranial kemia. Blood 1990; 76: 1449–1463. irradiation: A report from the Children’s Cancer Group. J Clin 5 Gaynon PS, Bleyer AW, Steinherz PG, Finklestein JZ, Littman P, Oncol 1998; 16: 920–930. Miller DR, Reaman G, Sather H, Hammond GD. Day 7 marrow 20 Richards S, Burrett J, Hann I, Chessells J, Hill F, Bailey C. Improved response and outcome for children with acute lymphoblastic leu- survival with early intensification: combined results from Medical kemia and unfavorable features. Med Pediatr Oncol 1990; 18: Research Council Childhood ALL randomized trials, UKALL X and 273–279. UKALL XI. Leukemia 1998; 12: 1031–1036. 6 Pui C-H, Behm FG, Crist WM. Clinical and biologic relevance 21 Tubergen DG, Gilchrist, O’Brien RT, Coccia PF, Sather HN, Was- of immunologic marker studies in childhood acute lymphoblastic kerwitz MJ, Hammond GD. Improved outcome with delayed leukemia. Blood 1993; 82: 343–362. intensification for children with acute lymphoblastic leukemia and 7 Camitta BM, Pullen J, Murphy S: Biology and treatment of acute intermediate presenting features: A Children’s Cancer Group lymphocytic leukemia in children. Sem Oncol 1997; 24: 83–91. Phase III Trial. J Clin Oncol 1993; 11: 527–537. 8 Shuster JJ, Camitta BM, Pullen, Borowitz MJ, Carroll AJ, Look AT, 22 Wheeler K. Chessells JM, Bailey CC, Richards SM. Treatment Mahoney DH, Mahmoud H, Lauer SJ, Land VJ. Identification of related deaths during induction and in first remission in acute lym- newly diagnosed children with acute lymphocytic leukemia at phoblastic leukemia: MRCUKALLX. Arch Dis Child 1996; 74: high risk for relapse. Cancer Res Ther and Control 1999; 9: 101–107. 101–107. 23 Goldie JH, Coldman AJ, Gudauskas GA: Rationale for the use of 9 Steinherz PG, Gaynon P, Miller DR, Reaman G, Bleyer A, Fin- alternating non- cross-resistant chemotherapy. Cancer Treat Rep klestein J, Evans RG, Meyers P, Steinherz LJ, Sather H, Hammond 1982; 66: 439–449. D. Improved disease-free survival of children with acute lym- 24 Borowitz MJ, Carroll AJ, Shuster JJ, Look AT, Behm FG, Pullen DJ, phoblastic leukemia at high risk for early relapse with the New Land VJ, Steuber P, Crist WM. Use of clinical and laboratory fea- York regimen – a new intensive therapy protocol: a report from tures to define prognostic subgroups in B-precursor acute lym- the Children’s Cancer Study Group. J Clin Oncol 1986; 4: 744–52. phoblastic leukemia: experience of the Pediatric Oncology Group. 10 Gaynon PS, Bleyer WA, Steinherz PG, Finklestein JZ, Littman PS, Rec Results Cancer Res. 1993; 131: 257–267. Miller DR, Reaman GH, Sather HN, Hammond GD. Modified 25 Trueworthy R, Shuster J, Look T, Crist W, Borowitz M, Carroll A, BFM therapy for children with previously untreated acute lym- Frankel L, Harris M, Wagner H, Haggard M, Mosijczuk A, Pullen phoblastic leukemia and unfavorable prognostic features; report of J, Steuber P, Land V. Ploidy of lymphoblasts is the strongest predic- the Children’s Cancer Study Group Study CCG-193P. Am J Pediatr tor of treatment outcome in B-precursor cell acute lymphoblastic Hematol Oncol 1988; 10: 42–50. leukemia. A Pediatric Oncology Group Study. J Clin Oncol 1992; 11 Rivera GK, Raimondi SA, Hancock ML, Behm FG, Pui CH, Abra- 10: 606–613. mowitch M, Mirro J Jr, Ochs JS, Look AT, Williams DL, Murphy 26 Pullen DJ, Crist WM, Falletta JM, Boyette JM, Roper M, Dowell SB, Dahl GV, Kalwinsky DK, Evans WE, Kun LE, Simone JV, Christ B, van Eys J, Humphrey GB, head D, Brock BL, Blackstock R, WM. Improved outcome in childhood acute lymphoblastic leuke- Metzgar RS, Cooper MD. A Pediatric Oncology Group classi- mia with reinforced early treatment and rotational combination fication protocol for acute lymphocytic leukemia (ALinC 13): chemotherapy. Lancet 1991; 337: 61–66. Immunologic phenotypes and correlation with treatment results. 12 Gaynon PS, Steinherz PG, Bleyer WA, Ablin AR, Albo VC, Fin- In Murphy SB, Gilbert JR (eds). Leukemia Research: Advances in klestein JZ, Grossman NJ, Novak LJ, Pyesmany AF, Reaman GH, Cell Biology and Treatment. Elsevier: Amsterdam, 1994, pp Chappell RJ, Sather HN, Hammond GD. Improved therapy for 221–239 children with acute lymphoblastic leukemia and unfavorable 27 Crist WM, Carroll A, Shuster JJ, Behm FG, Whitehead M, Vietti

Leukemia Comparing early intensive therapies in high-risk ALL SJ Lauer et al 1045 TJ, Look AT, Mahoney D, Ragab A, Pullen DJ, Land VJ. Poor prog- 35 Smith M, Arthur D, Camitta B, Carroll AJ, Crist W, Gaynon P, nosis of children with pre-B acute lymphoblastic leukemia with Gelber R, Heerema N, Korn EL, Link M, Murphy S, Pui CH, Pullen the t (1;19) (q23;p13): A Pediatric Oncology Group Study. Blood J, Reamon G, Sallan SE, Sather H, Shuster J, Simon R, Trigg M, 1990; 76: 117–122, Tubergen D, Uckun F, Ungerleider R. Uniform approach to risk 28 Fletcher JA, Lynch EA, Kimball VM, Donnelly M, Tantravahi R, classification and treatment assignment for children with acute Sallan SE. Translocation (9;22)is associated with extremely poor lymphoblastic leukemia. J Clin Oncol 1996; 14: 18–24. prognosis in intensively treated children with acute lymphoblastic 36 Harris MB, JJ Shuster, Pullen J, Borowitz MJ, Carroll AJ, Behm FG, leukemia. Blood 1991; 77: 435–439. Land VJ. Consolidation therapy with antimetabolite – based ther- 29 Shuster JJ. Handbook of Sample Size Guidelines for Clinical Trials. apy in standard risk acute lymphocytic leukemia of childhood: A CRC Press: Boca Raton, 1992. Pediatric Oncology Group Study. J Clin Oncol 1998; 16: 2840– 30 Kaplan EL, Meier P. Nonparametric estimation from incomplete 2847. observation. J Am Stat Assoc 1958; 53: 457–481. 37 Chessells JM, Bailey C, Richards SM: Intensification of treatment 31 Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, and survival in all children with lymphoblastic Leukemia. Results Mantel N, McPherson K, Peto J, Smith PG. Design and analysis of UR Medical Research Council trial UKALL X. Lancet 1995; 345: of randomized clinical trials requiring prolonged observation of 143–147. each patient: II. Analysis and examples. Br J Cancer 1977; 35: 38 Pui CH, Simone JV, Hancock ML, Evans WE, Williams DL, Bow- 1–39. man WP, Dahl GV, Dodge RK, Ochs J, Abromowitch M, Rivera 32 Cox DR. Regression models and life-tables. J R Stat Soc 1972; 34: GK. Impact of three methods of treatment intensification on acute 187–220. lymphoblastic leukemia in children. Long-term results of St. Jude’s 33 Mahoney DH, Shuster J, Nitschke R, Lauer SJ, Winick N, Steuber Total Therapy Study X. Leukemia 1992; 6: 150–157. CP, Camitta B. Intermediate-dose intravenous methotrexate with 39 Fojo AT, Ueda K, Slamon DJ, Poplack DG, Gottesman MM, Pastan intravenous mercaptopurine is superior to repetitive low-dose oral I. Expression of a multi-drug resistant gene in human tumors and methotrexate with intravenous mercaptopurine for children with tissues. Proc Natl Acad Sci USA 1987; 84: 265–269. lower-risk B-lineage acute lymphoblastic leukemia: A Pediatric 40 Bhalla K, Hindenburg A, Taub RN, Grant S. Isolation and charac- Oncology Group Phase III Trial. J Clin Oncol 1998; 16: 246–254. terization of an -resistant human leukemia cell line. 34 Mahoney DH, Shuster JJ, Nitschke R, Lauer SJ, Steuber CP, Winick Cancer Res 1985; 45: 3657–6662. N, Camitta B. Acute neurotoxicity in children with B-precursor 41 Kohn W. DNA topoisomerase as targets of anticancer drug action. acute lymphoid leukemia: An association with intermediate-dose Proc Am Assoc Cancer Res 1989; 30: 669–671. intravenous methotrexate and intrathecal triple therapy-A Pedi- atric Oncology Group Study. J Clin Oncol 1998; 16: 1712–1722.

Leukemia