Educational Report | Published:

Acute Leukemias

Long-term results of Dana-Farber Cancer Institute ALL Consortium protocols for children with newly diagnosed acute lymphoblastic leukemia (1985–2000)

Leukemia 24, 320–334 (2010) | Download Citation

Subjects

Abstract

The Dana-Farber Cancer Institute (DFCI) acute lymphoblastic leukemia (ALL) Consortium has been conducting multi-institutional clinical trials in childhood ALL since 1981. The treatment backbone has included 20–30 consecutive weeks of asparaginase during intensification and frequent vincristine/corticosteroid pulses during the continuation phase. Between 1985 and 2000, 1457 children aged 0–18 years were treated on four consecutive protocols: 85-01 (1985–1987), 87-01 (1987–1991), 91-01 (1991–1955) and 95-01 (1996–2000). The 10-year event-free survival (EFS)±s.e. by protocol was 77.9±2.8% (85-01), 74.2±2.3% (87-01), 80.8±2.1% (91-01) and 80.5±1.8% (95-01). Approximately 82% of patients treated in the 1980s and 88% treated in the 1990s were long-term survivors. Both EFS and overall survival (OS) rates were significantly higher for patients treated in the 1990s compared with the 1980s (P=0.05 and 0.01, respectively). On the two protocols conducted in the 1990s, EFS was 79–85% for T-cell ALL patients and 75–78% for adolescents (age 10–18 years). Results of randomized studies revealed that dexrazoxane prevented acute cardiac injury without adversely affecting EFS or OS in high-risk (HR) patients, and frequently dosed intrathecal chemotherapy was an effective substitute for cranial radiation in standard-risk (SR) patients. Current studies continue to focus on improving efficacy while minimizing acute and late toxicities.

Introduction

The Dana-Farber Cancer Institute (DFCI) began conducting randomized clinical trials on childhood acute lymphoblastic leukemia (ALL) in 1972. Studies performed during the 1970s showed improved event-free survival (EFS) for children who received doxorubicin (in addition to vincristine and prednisone) during the remission induction phase,1 and for those who received weekly Escherichia coli asparaginase during post-remission consolidation.2

In 1981, DFCI and several other institutions in the United States and Canada formed the DFCI ALL Consortium. The therapeutic backbone of Consortium trials included an intensive, multiagent induction phase, 20–30 weeks of asparaginase during post-remission consolidation and frequent vincristine/corticosteroid pulses during the continuation phase. On Protocol 81-01 (1981–1985), which was the first study conducted by the Consortium, the patients were stratified for the first time into two risk groups. Therapy was de-intensified for patients considered at lower risk of relapse based on age, leukocyte count and immunophenotype; such patients received lower cumulative doses of both anthracycline and corticosteroid. The overall EFS for patients enrolled in that study (74% at 5 years) was relatively favorable compared with contemporaneous childhood ALL trials, especially for children with T-cell ALL (5-year EFS 77%).3

Clinical trials conducted between 1985 and 2000 focused on improving survival rates while minimizing acute and late toxicities.4, 5, 6, 7 Strategies that were tested to improve survival included: substitution of dexamethasone for prednisone during post-induction therapy (Protocol 91-01),6 use of high-dose intravenous (i.v.) instead of standard-dose oral 6-mercaptopurine (6-MP) during the first of year of treatment (Protocol 91-01),6 addition of high-dose methotrexate during the remission induction phase5 and intensification of treatment for patients considered at very high risk of relapse, including patients with presenting leukocyte counts >100 × 109/l and infants (Protocols 85-01, 87-01 and 91-01).4, 5, 8 Attempts to reduce toxicity included: testing continuous infusion doxorubicin (Protocol 91-01) and the addition of a cardioprotectant, dexrazoxane (Protocol 95-01) in high-risk patients to minimize anthracycline-associated cardiotoxicity,6, 7 testing alternative preparations of asparaginase (Protocols 91-01 and 95-01)6, 7 and substituting intrathecal chemotherapy for cranial radiation in lower-risk patients (Protocols 87-01 and 95-01).5, 7 In this report, we update the results of the four consecutive clinical trials conducted by the DFCI ALL Consortium between 1985 and 2000.

Patients and methods

Between 1985 and 2000, 1457 children aged 0–18 years with newly diagnosed ALL (excluding mature B-cell ALL) were enrolled on four consecutive DFCI ALL Consortium protocols: 85-01 (1981–1985, N=220), 87-01 (1987–1991, N=369), 91-01 (1991–1995, N=377) and 95-01 (1996–2000, N=491). Patients were enrolled from the following DFCI ALL Consortium institutions: DFCI/Children's Hospital Boston (1985–2000), Hospital Sainte Justine, Montreal (1987–2000), Le Centre Hospitalier de L’Universite Laval, Quebec (1991–2000), Maine Medical Center/Maine Children's Cancer Program (1985–2000), McMaster Children's Hospital, Ontario (1985–2000), Mount Sinai Medical Center (1985–2000), Ochsner Clinic, New Orleans (1985–2000), San Jorge Children's Hospital, Puerto Rico (1991–2000), University of Massachusetts Medical Center (1985–1995), University of Puerto Rico, San Juan (1985–1991) and University of Rochester Medical Center, New York (1985–2000). The institutional review board of each participating institution approved all protocols. Informed consent was obtained from parents or guardians before instituting therapy.

Therapy

Details of therapy have been previously published.4, 5, 6, 7 Treatment was assigned based on risk group classification determined at the time of diagnosis (Table 1). On all protocols, there were four phases of therapy: Remission induction, central nervous system (CNS)-directed treatment, intensification and continuation. Details of each phase of therapy are shown in Table 2. On Protocols 85-01, 87-01 and 91-01, the remission induction phase was preceded by an investigational window, which consisted of a single agent given 3–5 days before the initiation of multiagent chemotherapy. Results of investigational windows have been previously reported.9, 10, 11

Table 1: Risk group classification on DFCI ALL Consortium Protocols (1985–2000)
Full size table
Table 2: Therapy on DFCI ALL Consortium Protocols: 1985–2000
Full size table

The major differences in therapy among the trials were as follows:

  • Asparaginase: On Protocols 85-01 and 87-01, patients received 20 weeks of intramuscular E. coli asparaginase during the intensification phase at a dose of 25 000 IU/m2/week. On Protocol 91-01, patients received 30 weeks of asparaginase during the intensification phase and were randomized to receive either E. coli asparaginase 25 000 IU/m2/week or polyethylene glycol (PEG)-asparaginase 2500 IU/m2 every 2 weeks. On Protocol 95-01, patients were randomized to receive either E. coli or Erwinia asparaginase 25 000 IU/m2/week for 20 weeks during the intensification phase.

  • Doxorubicin: The cumulative dose of doxorubicin for standard-risk (SR) patients on all protocols was 60 mg/m2. For high-risk (HR) patients, the cumulative dose of doxorubicin was 360 mg/m2 on Protocols 85-01, 87-01 and 91-01, and was 300 mg/m2 on Protocol 95-01. On Protocol 91-01, HR patients were randomized to receive doxorubicin 30 mg/m2/dose either as an i.v. bolus or a 48-h continuous infusion. On Protocol 95-01, HR patients were randomized to receive doxorubicin 30 mg/m2/dose i.v. bolus alone or immediately preceded by dexrazoxane 300 mg/m2/dose.

  • Corticosteroid: On Protocols 85-01, 87-01 and 95-01, prednisone was used during the intensification and continuation phases. On Protocol 91-01, dexamethasone was used instead of prednisone during these phases.

  • Methotrexate during induction: On Protocol 85-01, patients received low-dose methotrexate (40 mg/m2) as a single dose during the remission induction phase. On Protocol 87-01, patients were randomized to receive an induction dose of methotrexate as either low dose or high dose (4 gm/m2 over 1 h, followed by leucovorin rescue). On Protocols 91-01 and 95-01, a single dose of high-dose methotrexate was given during induction.

  • 6-MP: On Protocol 91-01, patients were randomized to receive either standard, oral 6-MP (50 mg/m2/day on days 1–14 every 3 weeks) or high-dose, i.v. 6-MP (1000 mg/m2/dose over 20 h weekly × 2 every 3 weeks) for 1 year after completion of the remission induction phase; thereafter, all patients received standard, oral 6-MP. On Protocols 85-01, 87-01 and 95-01, all patients received standard, oral 6-MP during all post-induction phases of treatment.

  • CNS-directed therapy:

    • ○ Protocol 85-01: SR patients received 18 Gy cranial radiation and HR patients received 24 Gy cranial radiation (22 Gy for patients aged 12–24 months). Cranial radiation was delayed until age 12 months for patients diagnosed during infancy. Total percentage of patients receiving cranial radiation was 100%.

    • ○ Protocol 87-01: All SR patients were initially treated without cranial radiation. Because of a higher-than-expected incidence of CNS relapse in SR boys, the protocol was amended in 1992 to allow 1 year of additional therapy for any SR boy in first remission, as previously described.5 Additional therapy included 18 Gy cranial radiation. Out of 60 eligible SR boys, 40 received this additional therapy. HR patients received 18 Gy cranial radiation in either daily or twice-daily fractions (randomized). Total percentage of patients receiving cranial radiation (including SR boys who received additional CNS-directed therapy after protocol amendment) was 66%.

    • ○ Protocol 91-01: SR girls were treated without radiation. SR boys and all HR patients received 18 Gy cranial radiation in either daily or twice-daily fractions (randomized). Total percentage of patients receiving cranial radiation was 76%.

    • ○ Protocol 95-01: SR patients were randomized to receive either intrathecal chemotherapy alone (every 9 weeks × 6 doses, and then every 18 weeks) without radiation or 18 Gy cranial radiation. SR girls who met Protocol 91-01 SR criteria were directly assigned to receive no radiation. HR patients received 18 Gy cranial radiation in either daily or twice-daily fractions (randomized). Total percentage of patients receiving cranial radiation was 60%.

  • Philadelphia chromosome; t(9;22): Beginning in 1989, patients with t(9;22) were treated with allogeneic transplantation in first remission. This was the only indication for transplantation in first remission on all studies conducted after that date. The percentage of patients who were transplanted in first remission by study was 0% (Protocol 85-01), <1% (Protocol 87-01), 2% (Protocol 91-01) and <1% (Protocol 95-01).

  • Investigational Window: On Protocol 85-01, patients received a single dose of E. coli asparaginase, randomized to either 25 000 IU/m2 or 2500 IU/m2, administered 5 days before the initiation of the remission induction phase. On Protocol 87-01, patients received a single dose of asparaginase, randomized to E. coli 25 000 IU/m2, Erwinia 25 000 IU/m2 or PEG 2500 IU/m2, given 5 days before the initiation of the remission induction phase. On Protocol 91-01, patients received 3 days of corticosteroids, immediately followed by the remission induction phase. Those patients were randomized to receive prednisone 40 mg/m2/day or dexamethasone 6, 18 or 150 mg/m2/day for 3 days. Protocol 95-01 did not have an Investigational Window.

Statistical analysis

Outcome events included induction failure, induction death, death during remission, relapse and second malignancy (meningiomas, basal cell carcinomas and benign tumors were not considered second malignancies). For Protocols 85-01, 87-01 and 91-01, induction failure was defined as the failure to achieve complete remission (CR) at day 52 after diagnosis. For Protocol 95-01, induction failure was defined as persistent leukemia at day 30 after diagnosis. EFS was measured from the date of complete remission to the first event or until the date of last contact for event-free survivors. For EFS, induction failure and induction death were considered events at time zero. Overall survival (OS) was measured from the date of starting treatment to death from any cause. EFS and OS were estimated using the Kaplan–Meier method and compared with the log-rank test.12 Multivariable regression was performed using the Cox's proportional hazard model to assess prognostic factors for EFS and OS for each protocol.13

Cumulative incidence functions of any CNS relapse, isolated CNS (no other site involved) and any testicular relapse were constructed using the method of Kalbfleish and Prentice14 and compared with Gray's test15 for patients who achieved CR. In the estimation of these functions, all other failures were considered competing events.

Results

Between 1985 and 2000, 1457 children were enrolled in DFCI ALL Consortium Protocols. Tables 3 and 4 summarize the outcomes by protocol. Of the 1423 patients who achieved CR, 241 patients (17%) relapsed. All but three relapses occurred within 7.4 years from diagnosis. A total of 30 patients (2.1%) died in first remission and eight patients (0.6%) were diagnosed with a second malignant neoplasm (SMN) as their first event. As of December 2008, 1246 patients (86%) are alive, of whom 1144 have never relapsed or been diagnosed with an SMN. Both the EFS and OS improved significantly from the clinical trials conducted in the 1980s (Protocol 85-01 and 87-01) to those in the 1990s (Protocols 91-01 and 95-01) (P=0.05 and 0.01, respectively; Figures 1 and 2).

Table 3: Outcome by protocol (1985–2000)
Full size table
Table 4: Timing of events for patients achieving a CR
Full size table
Figure 1
Figure 1

Event-free survival (EFS) by decade. The EFS of protocols conducted in the 1990s (91-01 and 95-01) was superior to that of protocols conducted in the 1980s (85-01 and 87-01), P=0.05.

Full size image
Figure 2
Figure 2

Overall survival (OS) by decade. The OS of protocols conducted in the 1990s (91-01 and 95-01) was superior to that of protocols conducted in the 1980s (85-01 and 87-01), P=0.01.

Full size image

Protocol 85-01(1985–1987)

Outcome on Protocol 85-01 is summarized in Table 3 and Figure 3. Median follow-up was 13.8 years. Of the 220 evaluable patients, 217 entered CR (99%), 37 relapsed (16.8%) and 8 patients died in CR (3.6%). In all, 171 (78%) remain alive and free of adverse events. The 10-year cumulative incidence (CI) estimates for isolated marrow and any marrow relapses were 12.2±2.2 and 13.1±2.3%, respectively. The 10-year CI estimates for isolated CNS and any CNS relapses were 2.8±1.1 and 3.7±1.3%, respectively. Of the 116 evaluable male patients, the 10-year CI of any testicular relapse was 0.9±0.9% (1 patient had an isolated testicular relapse). One patient experienced an SMN (acute myelogenous leukemia) as a first event. One other patient developed a basal cell carcinoma in a previous radiation field. Two patients (1.0%) experienced their first event after 5 years of complete CR (1 relapse and 1 remission death; Table 4). The 10-year EFS and OS were 77.9±2.8 and 80.9±2.7%, respectively. For SR patients, the 10-year EFS and OS rates were 88.8±3.5 and 92.4±3.0%, and the EFS and OS rates for HR and very high-risk (VHR) patients were 71.6±3.9 and 74.2±3.8%, respectively. Univariate predictors of outcome are shown in Table 5. Multivariable regression analysis, including age, sex, presenting leukocyte count, phenotype and CNS status at diagnosis, identified only presenting white blood cells (WBCs) 100 K as an adverse independent predictor of both EFS (hazard ratio 5.08, P<0.01) and OS (hazard ratio 5.71, P<0.01).

Figure 3
Figure 3

Event-free survival and cumulative incidence of isolated or any central nervous system (CNS) relapse for 220 patients treated in Protocol 85-01 (1985–1987). Median follow-up was 13.8 years.

Full size image
Table 5: Protocol 85-01 outcome by patient characteristics
Full size table

Protocol 87-01 (1987–1991)

Outcome on Protocol 87-01 is summarized in Table 3 and Figure 4. Median follow-up was 13.3 years. Of the 369 evaluable patients, 356 entered CR (96%), 72 relapsed (19.5%) and 7 patients died in CR (1.9%). In all, 274 (74%) remain alive and free of adverse events. The 10-year CI estimates for isolated marrow and any marrow relapses were 13.1±1.8 and 15.9±2.0%, respectively. The 10-year CI estimates for isolated CNS and any CNS relapses were 4.2±1.1 and 5.9±1.3%, respectively. Of the 216 evaluable male patients, the 10-year CI of any testicular relapse was 1.0±0.7% (no isolated testicular relapses observed). Three patients experienced an SMN as a first event (two cases of acute myelogenous leukemia and one parotid gland carcinoma). The parotid gland had been previously irradiated as part of the CNS-directed therapy. Other tumors arising within previous fields of radiation included a meningioma in one patient and a benign fibrous tumor of the left orbit in another patient. A total of 10 patients (2.7%) experienced their first event after 5 years of complete CR (seven relapses, one SMN and two remission deaths; Table 4). The 10-year EFS and OS were 74.2±2.3 and 83.3±2.0%, respectively. For SR patients, the 10-year EFS and OS rates were 77.4±3.5 and 92.1±2.3%, and the rates for HR/VHR patients were 72.2±3.0 and 77.8±2.8%.

Figure 4
Figure 4

Event-free survival and cumulative incidence of isolated or any central nervous system (CNS) relapse for 369 patients treated in Protocol 87-01 (1987–1991). Median follow-up was 13.3 years.

Full size image

Outcome by patient characteristics is presented in Table 6. On univariate analysis, male sex, in addition to age and presenting leukocyte count, was identified as a significant predictor of both EFS and OS, primarily because of a high rate of CNS relapses in SR boys compared with SR girls (10-year CI of 20.3±4.6% for boys compared with 4.8±2.7% for girls, P<0.01). Multivariable regression analysis, including age, sex, presenting leukocyte count, phenotype and CNS status at diagnosis, identified male sex (hazard ratio 1.79, P=0.01), WBCs 10–49 K (hazard ratio 2.33, P<0.01) and WBCs 100 K (hazard ratio 2.69, P<0.01) as independent adverse predictors of EFS. WBCs 100 K (hazard ratio 3.04, P<0.01), WBCs 10–49 K (hazard ratio 2.22, P<0.01) and age 10 years (hazard ratio 2.11, P<0.01), but not male sex, were independent adverse predictors of OS.

Table 6: Protocol 87-01 outcome by patient characteristics
Full size table

Protocol 91-01 (1991–1996)

Outcome on Protocol 91-01 is summarized in Table 3 and Figure 5. Median follow-up was 12.5 years. Of the 377 evaluable patients, 370 entered CR (98%), 53 relapsed (14.1%) and 12 patients died in CR (3.2%). In all, 304 (81%) remain alive and free of adverse events. The 10-year CI estimates for isolated marrow and any marrow relapses were 8.8±1.5 and 12.2±1.7%, respectively. The 10-year CI estimates for isolated CNS and any CNS relapses were 1.1±0.5 and 4.2±1.1%, respectively. Of the 199 evaluable male patients, the 10-year cumulative incidence for isolated or any testicular relapse was 1.0±0.7% and 1.5±0.9%. One patient experienced an SMN as a first event (malignant brain tumor in a previously irradiated patient) and two others were diagnosed with meningiomas (both previously irradiated). A total of 11 patients (3.0%) experienced their first event after 5 years of complete CR (10 relapses and 1 SMN; Table 4). The 10-year EFS and OS were 80.8±2.1 and 86.2±1.8%, respectively. For SR patients, the 10-year EFS and OS rates were 84.3±3.2 and 91.0±2.5, and the rates for HR/VHR patients were 78.9±2.7 and 83.5±2.4%, respectively.

Figure 5
Figure 5

Event-free survival and cumulative incidence of isolated or any central nervous system (CNS) relapse for 377 patients treated in Protocol 91-01 (1991–1995). Median follow-up was 12.5 years.

Full size image

Outcome by patient characteristic is presented in Table 7. Multivariable regression analysis, including age, sex, presenting leukocyte count, phenotype and CNS status at diagnosis, identified only CNS-3 status as an independent adverse predictor of EFS (hazard ratio 4.36, P=0.02). Although no independent predictors were identified when these same variables were included in a multivariable regression analysis for OS, univariate analysis indicated that HR/VHR patients had a significantly lower OS than SR patients (P=0.04).

Table 7: Protocol 91-01 outcome by patient characteristics
Full size table

Protocol 95-01 (1996–2000)

Outcome on Protocol 95-01 is summarized in Table 3 and Figure 6. Median follow-up was 8.6 years. Of the 491 evaluable patients, 480 entered CR (98%), 79 relapsed (16%) and 3 patients died in CR (0.6%). In all, 395 (80%) remain alive and free of adverse events. The 10-year CI estimates for isolated marrow and any marrow relapses were 12.1±1.5 and 15.9±1.8%, respectively. The 10-year CI estimates for isolated CNS and any CNS relapses were 0.7±0.4 and 3.8±1.0%, respectively. Of the 274 evaluable male patients, the 10-year CI of any and isolated testicular relapse was 1.9±0.9 and 0.8±0.5%, respectively. Three patients experienced an SMN as a first event (one malignant brain tumor in a previously irradiated patient and two cases of malignant melanoma). Neither case of malignant melanoma occurred in a previous radiation field. A total of eight patients (1.7%) experienced their first event after 5 years of CCR (six relapses and two SMN; Table 4). All six of these very late relapses occurred in SR patients. The 10-year EFS and OS were 79.0±2.1 and 88.9±1.5%, respectively. For SR patients, the 10-year EFS and OS rates were 83.1±2.5 and 93.1±2.1%, and the rates for HR/VHR patients were 74.1±3.3 and 83.7±2.5%, respectively.

Figure 6
Figure 6

Event-free survival and cumulative incidence of isolated or any central nervous system (CNS) relapse for 491 patients treated in Protocol 95-01 (1996–2000). Median follow-up was 8.6 years.

Full size image

Univariate predictors of outcome are showed in Table 8. End-induction (day 30) minimal residual disease (MRD) level was measured using PCR in 284 of 430 (66%) B-precursor patients who achieved morphologic CR16 and was a significant predictor of both EFS and OS. The 10-year EFS for B-precursor patients with low MRD (<0.001) was 83.9±3.0 versus 24.4±7.1% for those with high MRD (0.001), P<0.01. The presence or absence of the TEL/AML1 fusion was also prospectively tested using PCR in 299 of 438 (68%) patients with B-precursor phenotype.17 The TEL/AML1 fusion was detected in 26% of these patients, and was associated with significantly better OS (P=0.05), but not EFS (P=0.10). Multivariable regression analysis, including age, sex, presenting leukocyte count, phenotype and CNS status at diagnosis, identified T-cell phenotype as an independent favorable predictor of EFS (hazard ratio 0.39, P=0.02) and OS (hazard ratio 0.36, P=0.04). Independent adverse predictors of EFS included WBCs 100 K (hazard ratio 3.40, P<0.01) and age 10 years (hazard ratio 1.66, P=0.04). These two features were also independent adverse predictors of OS (WBCs 100 K: hazard ratio 5.10, P<0.01; and age 10 years: hazard ratio 2.90, P<0.01).

Table 8: Protocol 95-01 outcome by patient characteristics
Full size table

Outcome of randomized comparisons

Asparaginase

On Protocol 91-01, 198 patients (SR and HR/VHR) were randomized to receive either native E. coli asparaginase (25 000 IU/m2 IM weekly) or PEG-asparaginase (2500 IU/m2 intramuscular every 2 weeks) for a total of 30 weeks during post-induction consolidation. There was no significant difference in EFS (P=0.29) or OS (P=0.29) based on asparaginase type (Table 7). On Protocol 95-01, 286 patients (SR and HR/VHR) were randomized to receive either native E. coli or Erwinia asparaginase (both dosed at 25 000 IU/m2 intramuscular weekly) for 20 weeks during post-induction consolidation. Patients randomized to receive Erwinia asparaginase had a significantly inferior 10-year EFS (75.2±3.8 vs 84.6±3.4%, P=0.02) and OS (85.3±3.1 vs 93.1±2.1%, P=0.04 Table 8). More patients randomized to Erwinia experienced a relapse involving the CNS (7 vs 1%, P<0.01).

Doxorubicin

On Protocol 91-01, 204 HR/VHR patients were randomized to receive doxorubicin (30 mg/m2) as either a bolus dose or a 48-h continuous infusion every 3 weeks to a total cumulative dose of 360 mg/m2 during the post-induction consolidation phase. There was no difference in EFS (P=0.24) or OS (P=0.31) based on infusion duration (Table 7). On Protocol 95-01, 205 HR/VHR patients were randomized to receive bolus doxorubicin (30 mg/m2) every 3 weeks with or without dexrazoxane (300 mg/m2), a potential cardioprotectant agent. Total cumulative dose of doxorubicin was 300 mg/m2. There was no difference in EFS (P=0.81) or OS (P=0.66) when comparing patients treated with or without dexrazoxane (Table 8). No SMNs have been observed in patients randomized to receive dexrazoxane.

Central nervous system (CNS)-directed therapy

Standard Risk: On Protocol 95-01, 164 SR patients were randomized to receive either frequently dosed triple IT chemotherapy (methotrexate/cytarabine/hydrocortisone) without radiation or 18 Gy cranial radiation with less frequent IT therapy. There was no difference in EFS (P=0.21), OS (P=0.39) or CI of isolated CNS relapse (P=0.15) between the two randomized groups (Table 8).

High Risk: On Protocols 87-01, 91-01 and 95-01, HR/VHR patients were randomized to receive either daily (180 cGy) or twice-daily (90 cGy) fractions of cranial radiation to a total dose of 18 Gy. A total of 591 participated in these randomizations. There was no difference in 10-year EFS (P=0.47), OS (P=0.59), CI of isolated CNS relapse (P=0.18) or CI of any CNS relapse (P=0.13).

Other Randomizations

On Protocol 87-01, 353 patients (SR and HR/VHR) were randomized to receive either high-dose (4 gm/m2) or low-dose (40 mg/m2) methotrexate during remission induction. There was no difference in EFS (P=0.62) or OS (P=0.66) based on methotrexate dose (Table 6). On Protocol 91-01, 322 patients (SR and HR/VHR) were randomized to receive high-dose, i.v. 6-MP or standard, low-dose oral 6-MP during the first year of post-induction therapy. There was no difference in EFS (P=0.99) or OS (P=0.66) based on 6-MP dosing (Table 7).

Discussion

On the four consecutive DFCI ALL Consortium protocols conducted between 1985 and 2000, we observed long-term EFS rates ranging from 74 to 81% and OS rates from 81 to 89%. Both EFS and OS rates significantly improved during the 1990s, with EFS rates exceeding 80% and OS approaching 90% for patients treated during that decade. Although the incidence of marrow-involved relapses was relatively unchanged over the 15-year period, the incidence of isolated CNS relapse decreased, likely contributing to the improvement in EFS in the 1990s. There was also a decrease in the remission death rate, likely secondary to improvements in supportive care. The improvements in OS may have also been due, in part, to improved salvage after relapse during the 1990s, as evidenced by a larger difference between EFS and OS rates for patients treated in the 1990s compared with the 1980s.

The improvement in outcome during the 1990s occurred at the same time that therapy was de-intensified on DFCI ALL Consortium protocols: during that decade, cumulative dosage of doxorubicin was decreased in higher-risk patients and fewer patients received cranial radiation. In addition, risk group definitions were changed during the 1990s that resulted in more patients being classified as SR and and hence receiving less intensive therapy than they would have during the 1980s (when they would have been considered HR).

The favorable outcomes on our trials are especially notable for subsets of HR/VHR patients who historically have had worse prognoses, including those with T-cell immunophenotype and adolescents (age 10–18 years at diagnosis). We have previously reported that patients with T-cell ALL treated in our protocols have similar outcomes to those with B-precursor phenotype.18 In fact, on Protocol 95-01, T-cell phenotype was an independent predictor of favorable EFS and OS on multivariable analysis. We have also previously shown that there was no significant difference in outcome between younger (10–14 years old) and older (15–18 years old) adolescents treated in our trials in the 1990s, with long-term EFS exceeding 75% for both subgroups.19 On the basis of the favorable outcomes achieved by older adolescents in our protocols, we are currently piloting our HR regimen in adults with ALL, with promising preliminary results.20

Although age and phenotype no longer identify patients at the highest risk of relapse, very high presenting leukocyte count 100 K remained an independent predictor of adverse outcome throughout this time period. Overall, the prognosis for such patients improved in the 1990s (EFS 66–70% vs 52–62% in the 1980s), perhaps because of some changes in the regimen backbone during this decade. For instance, outcomes for patients with a very high presenting leukocyte count were best on Protocol 91-01 (1991–1995), which included a more prolonged asparaginase consolidation phase (30 weeks instead of 20 weeks) and the use of dexamethasone instead of prednisone during all post-induction phases. In fact, Protocol 91-01 was the only trial reported in this study in which leukocyte count was not a significant prognostic factor.

On Protocol 95-01 (1996–2000), we identified end-induction MRD level as a significant independent predictor of outcome. Patients with high end-induction MRD (0.001 as measured using quantitative PCR) had a 10.5-fold greater risk of relapse than those with low MRD.16 On the basis of these results, we have re-defined the VHR group in our current clinical trial to include patients with high end-induction MRD, as well as those with the following adverse chromosomal abnormalities regardless of MRD level: mixed lineage leukemia (MLL) gene leukemia translocation, hypodiploidy (<45 chromosomes) and t(9;22). Approximately 15% of patients are now considered VHR and receive intensified treatment.

A major component of our therapeutic backbone is the administration of asparaginase for 20–30 consecutive weeks, beginning 3 weeks after the completion of the remission induction phase. Toxicities associated with this treatment have included hypersensitivity reactions in 20–30% of patients, pancreatitis in 5–8% and thrombotic events in 2–5%.6, 7 After allergy to native E. coli asparaginase, patients have been treated with alternative asparaginase preparations (either weekly PEG or twice-weekly Erwinia, depending on protocol and agent availability during this era); approximately one-third of patients develop hypersensitivity to the second asparaginase preparation. The incidence of pancreatitis and thromboembolic complications, but not asparaginase allergy, is higher in patients 10–18 years of age compared with those <10 years.19 In an attempt to optimize asparaginase dosing, we have extensively studied toxicities associated with the different asparaginase preparations. On Protocol 95-01, weekly Erwinia asparaginase was associated with a lower incidence of asparaginase-associated toxicity (10 vs 24%), but also with inferior 5-year EFS compared with weekly E. coli asparaginase.7 This result was likely due to the dosing schedule; because Erwinia asparaginase has a far shorter half-life than E. coli asparaginase, it is probable that fewer Erwinia-treated patients experienced continuous asparagine depletion during the intensification phase. On Protocol 91-01, we showed that intramuscular PEG-asparaginase (administered every 2 weeks) was associated with a reduced risk of hypersensitivity compared with weekly E. coli asparaginase without affecting EFS.6 However, that study was not sufficiently powered to detect small differences in EFS. In our current trial, we are comparing i.v. PEG-asparaginase with native E. coli asparaginase in a larger cohort of patients to determine the tolerability of i.v. administration of PEG-asparaginase, as well as the relative toxicity and efficacy of the two preparations.

A major focus of our clinical trials has been to reduce late effects of therapy. To that end, our therapeutic backbone does not include exposure to alkylating agents or epipodophyllotoxins. To minimize the risk of late cardiotoxicity, SR patients receive only 60 mg/m2 cumulative dose of doxorubicin. For HR patients, the cumulative dosage of doxorubicin was reduced from 360 to 300 mg/m2 in 1996. In successive randomized trials, we found that continuous infusion doxorubicin was not cardioprotective,21 but showed that dexrazoxane prevented acute cardiac injury (as measured by troponin-T elevation) in HR patients without increasing the risk of relapse or second malignant neoplasm.22, 23 We are currently analyzing long-term echocardiograms (obtained 5 or more years after completion of anthracycline) on patients who participated in that randomization, and continue to use dexrazoxane before each dose of doxorubicin in HR/VHR patients.

Between 1985 and 2000, we also focused on reducing late effects associated with CNS-directed therapy. On Protocol 87-01, all SR patients were treated without cranial radiation; however, no change was made in either intrathecal or systemic chemotherapy to substitute for the absence of radiation. This non-randomized change resulted in an unacceptably high rate of CNS relapses in SR boys, although most of these patients could be salvaged after relapse.5 On Protocol 95-01, we were able to successfully eliminate cranial radiation in all SR patients (including boys) by more frequent administration of intrathecal chemotherapy (every 9 weeks) during the first year of treatment.7 Neurocognitive testing of survivors from that study (median follow-up 6 years from diagnosis) showed that cognitive function for both irradiated and non-irradiated patients was solidly in the normal range, although irradiated patients as a group showed a slower rate of information processing.24 Longer follow-up is necessary to assess more fully the relative long-term neurocognitive and neuroendocrine consequences of these two CNS-directed treatments (with and without radiation). In addition to neurocognitive sequelae, cranial radiation has also been associated with a higher risk of SMNs.25 On our current trial, we have restricted the use of radiation to those considered to be at the highest risk of CNS relapse (∼25–30% of patients), including those with CNS-3 status at diagnosis, T-cell phenotype and/or high end-induction MRD. In an attempt to reduce the risk of radiation-associated late effects for these patients, we are also using a lower dose of cranial radiation (12 Gy instead of 18–24 Gy in earlier studies).

The EFS and OS of the two protocols that we conducted in the 1990s were nearly identical. The plateau in survival rates over that decade suggests that we may have reached the limits of currently applied risk factors and conventional chemotherapeutic agents. To improve outcomes, our current studies focus on identifying biologic factors that may supplement or replace the epidemiologic factors currently used to determine risk-based therapy. For instance, microarray gene expression studies from our investigators have identified biologically distinct and prognostically relevant subtypes of ALL based upon gene expression profiles.26, 27 In addition, research focused on pharmacogenomics has begun to identify patient-related factors that affect outcome and help lead to more individualized therapy.28 We have also focused on identifying novel, targeted therapies, including inhibitors of the fms-related tyrosine kinase 3 (FLT3),29 the antiapoptotic protein B-cell leukemia/lymphoma 2 (BCL-2)30 and the mammalian target of rapamycin (mTOR) pathway (which has been implicated in glucocorticoid resistance).31 By identifying new, biologically distinctive patient subsets and devising novel targeted treatments for them, our aim is to improve survival and reduce toxicities for all patients with ALL.

Conflict of interest

Dr Silverman received compensation as an advisory board member for EUSA Pharma, and also received compensation as a consultant for Enzon, Inc. Dr Sallan received honoraria and research funding from Enzon Inc. and compensation as an advisory board member for EUSA Pharma. Jane O’Brien, Kristen Stevenson, Eileen Whyte O’Holleran and Drs Asselin, Barr, Clavell, Cohen, Cole, Kelly, Laverdiere, Michon, Neuberg, Schorin and Schwartz declare no conflict of interest.

References

  1. 1.

    , , , , . Intermittent combination chemotherapy with adriamycin for childhood acute lymphoblastic leukemia: clinical results. Blood 1978; 51: 425–433.

  2. 2.

    , , , , , . Influence of intensive asparaginase in the treatment of childhood non-T- cell acute lymphoblastic leukemia. Cancer Res 1983; 43: 5601–5607.

  3. 3.

    , , , , , et al. Four-agent induction and intensive asparaginase therapy for treatment of childhood acute lymphoblastic leukemia. N Engl J Med 1986; 315: 657–663.

  4. 4.

    , , , , , et al. Treatment of childhood acute lymphoblastic leukemia: results of Dana- Farber Cancer Institute/Children's Hospital Acute Lymphoblastic Leukemia Consortium Protocol 85-01. J Clin Oncol 1994; 12: 740–747.

  5. 5.

    , , , , , et al. Treatment of childhood acute lymphoblastic leukemia: results of Dana- Farber ALL Consortium Protocol 87-01. J Clin Oncol 2002; 20: 237–246.

  6. 6.

    , , , , , et al. Improved outcome for children with acute lymphoblastic leukemia: results of Dana-Farber Consortium Protocol 91-01. Blood 2001; 97: 1211–1218.

  7. 7.

    , , , , , et al. Results of the Dana-Farber Cancer Institute ALL Consortium Protocol 95-01 for children with acute lymphoblastic leukemia. Blood 2007; 109: 896–904.

  8. 8.

    , , , , , et al. Intensified therapy for infants with acute lymphoblastic leukemia: results from the Dana-Farber Cancer Institute Consortium. Cancer 1997; 80: 2285–2295.

  9. 9.

    , , , , , . Comparative pharmacokinetic studies of three asparaginase preparations. J Clin Oncol 1993; 11: 1780–1786.

  10. 10.

    , , , , , et al. Prognostic significance of early response to a single dose of asparaginase in childhood acute lymphoblastic leukemia. J Pediatr Hematol Oncol 1999; 21: 6–12.

  11. 11.

    , , , , , et al. Improved response with higher corticosteroid dose in children with acute lymphoblastic leukemia. J Clin Oncol 2001; 19: 1040–1046.

  12. 12.

    , . Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481.

  13. 13.

    , . Regression models and life tables. J Royal Stat Soc 1972; B34: 187–220.

  14. 14.

    , . The Statistical Analysis of Failure Time Data. Wiley: New York, 1980.

  15. 15.

    . A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 1988; 16: 1141–1154.

  16. 16.

    , , , , , et al. Quantitative analysis of minimal residual disease predicts relapse in children with B-lineage acute lymphoblastic leukemia in DFCI ALL Consortium Protocol 95-01. Blood 2007; 110: 1607–1611.

  17. 17.

    , , , , , et al. Prospective analysis of TEL/AML1 positive patients treated on Dana-Farber Cancer Institute Consortium Protocol 95-01. Blood 2006; 107: 4508–4513.

  18. 18.

    , , , , , et al. Childhood T-cell acute lymphoblastic leukemia: the Dana-Farber Cancer Institute acute lymphoblastic leukemia consortium experience. J Clin Oncol 2003; 21: 3616–3622.

  19. 19.

    , , , , , et al. Favorable outcome for adolescents with acute lymphoblastic leukemia treated on Dana-Farber Cancer Institute Acute Lymphoblastic Leukemia Consortium Protocols. J Clin Oncol 2007; 25: 813–819.

  20. 20.

    , , , , , et al. A multicenter Phase II study using a dose intensified pediatric regimen in adults with untreated acute lymphoblastic leukemia. Blood 2007; 110: 181a.

  21. 21.

    , , , , , et al. Doxorubicin administration by continuous infusion is not cardioprotective: the Dana-Farber 91-01 Acute Lymphoblastic Leukemia protocol. J Clin Oncol 2002; 20: 1677–1682.

  22. 22.

    , , , , , et al. Dexrazoxane reduces myocardial injury in doxorubicin-treated children with acute lymphoblastic leukemia: results from a randomized trial. N Engl J Med 2004; 351: 145–153.

  23. 23.

    , , , , , et al. Absence of secondary malignant neoplasms in children with high-risk acute lymphoblastic leukemia treated with dexrazoxane. J Clin Oncol 2008; 26: 1106–1111.

  24. 24.

    , , , , , et al. Neuropsychological outcomes from a randomized trial of triple intrathecal chemotherapy compared with 18 Gy cranial radiation as CNS treatment in acute lymphoblastic leukemia: findings from Dana-Farber Cancer Institute ALL Consortium Protocol 95-01. J Clin Oncol 2007; 25: 4914–4921.

  25. 25.

    , , , , , et al. Cumulative incidence of secondary neoplasms as a first event after childhood acute lymphoblastic leukemia. JAMA 2007; 297: 1207–1215.

  26. 26.

    , , , , , et al. MLL translocations specify a distinct gene expression profile that distinguishes a unique leukemia. Nat Genet 2002; 30: 41–47.

  27. 27.

    , , , , , et al. Gene expression signatures define novel oncogenic pathways in T cell acute lymphoblastic leukemia. Cancer Cell 2002; 1: 75–87.

  28. 28.

    , , , , , et al. DNA variants in the dihydrofolate reductase gene and outcome in childhood ALL. Blood 2008; 111: 3692–3700.

  29. 29.

    , , , , , et al. Inhibition of FLT3 in MLL Validation of a therapeutic target identified by gene expression based classification. Cancer Cell 2003; 3: 173–183.

  30. 30.

    , , , , . BCL-2 dependence and ABT-737 sensitivity in acute lymphoblastic leukemia. Blood 2008; 111: 2300–2309.

  31. 31.

    , , , , , et al. Gene expression-based chemical genomics identifies rapamycin as a modulator of MCL1 and glucocorticoid resistance. Cancer Cell 2006; 10: 331–342.

Download references

Acknowledgements

These trials were supported in part by a grant from the National Institute of Health (NCI Grant 5P01CA068484). We thank the patients, families, physicians, nurses, clinical research coordinators and all others who participated in these trials. We acknowledge the important contributions of Jennifer Cronin, Annette Dalton, Virginia Dalton, Meghan Eaton and Richard D Gelber.

Author information

Affiliations

  1. Departments of Pediatric Oncology, Biostatistics, and Computational Biology, Dana-Farber Cancer Institute, Boston, MA, USA

    • L B Silverman
    • , K E Stevenson
    • , J E O'Brien
    • , E W O'Holleran
    • , D S Neuberg
    •  & S E Sallan

  2. Division of Hematology/Oncology, Children's Hospital Boston, Boston, MA, USA

    • L B Silverman
    •  & S E Sallan

  3. Division of Pediatric Hematology-Oncology, University of Rochester Medical Center, Rochester, NY, USA

    • B L Asselin

  4. Division of Pediatric Hematology/Oncology, McMaster University, Hamilton, Ontario, Canada

    • R D Barr

  5. Division of Pediatric Hematology/Oncology, San Jorge Children's Hospital, San Juan, PR, USA

    • L Clavell

  6. Division of Pediatric Hematology/Oncology, Montefiore Medical Center, New York, NY, USA

    • P D Cole

  7. Division of Pediatric Oncology, Columbia University Medical Center, Morgan Stanley Children's Hospital of New York-Presbyterian, New York, NY, USA

    • K M Kelly

  8. Division of Hematology and Oncology, Hospital Sainte Justine, University of Montreal, Montreal, Quebec, Canada

    • C Laverdiere

  9. Division of Hematology-Oncology, Centre Hospitalier U. de Quebec, Quebec City, Quebec, Canada

    • B Michon

  10. Inova Fairfax Hospital, Falls Church, VA, USA

    • M A Schorin

  11. Division of Pediatric Hematology-Oncology, Hasbro Children's Hospital, Warren Alpert Medical School of Brown University, Providence, RI, USA

    • C L Schwartz

  12. Department of Pediatrics, Stanford University School of Medicine, Palo Alto, CA, USA

    • H J Cohen

Authors

  1. Search for L B Silverman in:

  2. Search for K E Stevenson in:

  3. Search for J E O'Brien in:

  4. Search for B L Asselin in:

  5. Search for R D Barr in:

  6. Search for L Clavell in:

  7. Search for P D Cole in:

  8. Search for K M Kelly in:

  9. Search for C Laverdiere in:

  10. Search for B Michon in:

  11. Search for M A Schorin in:

  12. Search for C L Schwartz in:

  13. Search for E W O'Holleran in:

  14. Search for D S Neuberg in:

  15. Search for H J Cohen in:

  16. Search for S E Sallan in:

Corresponding author

Correspondence to L B Silverman.

To obtain permission to re-use content from this article visit RightsLink.

About this article

Publication history

Received

Accepted

Published

DOI

https://doi.org/10.1038/leu.2009.253