The Effects of Postinduction Intensification Treatment With Cytarabine and Daunorubicin in Adult Acute Lymphocytic Leukemia: A Prospective Randomized Clinical Trial by Cancer and Leukemia Group B By Rose Ruth Ellison, Rosemarie Mick, Janet Cuttner, Charles A. Schiffer, Richard T. Silver, Edward S. Henderson, Thomas Woliver, Ivor Royston, Frederick R. Davey, Arvin S. Glicksman, Clara D. Bloomfield, and James F. Holland Cancer and Leukemia Group B undertook a randomized trial of intensification treatment in adults aged 15 to 79 years with acute lymphocytic leukemia (ALL) in complete remission (CR). Daunorubicin (DNR), prednisone, vincristine (VCR), intrathecal (IT) methotrexate (MTX), and asparaginase produced 177 CRs in 277 patients. One hundred fifty-one patients were randomly assigned to receive treatment as follows: 74 received intensive cytarabine and DNR, and 77 received cycles of mercaptopurine (6-MP) and MTX, followed by 6MP, MTX, VCR, and prednisone for 3 years in all. One hundred twelve patients received CNS prophylaxis. Intensification produced major myelosuppression but did not improve remission duration (median, 21 months). Of the 151 patients with CRs who entered the intensification phase, 29% remain in continuous CR (43 to 117 months); in 19 patients, CRs have lasted for longer than 7 years. No relapses occurred after 60 months. Median survival from the

time of randomization was 30 months. Those under 30 years of age responded more frequently, with longer CR and survival. While 53% of those aged 15 to 19 years remain in continuous CR, 92% of patients over 59 years have relapsed. The presence of a myeloid antigen on the leukemic cells was adversely prognostic for CR achievement and for survival. Pretreatment WBC and platelet levels independently affected CR duration and survival. Early MI marrow development presaged longer remissions. CNS relapse occurred in 47 of 256 patients with normal CSF before treatment, in 29 before CNS prophylaxis. CNS disease occurred after CNS prophylaxis in 18 patients: 13 of 61 who had received standard premaintenance and five of 51 who received intensification. No advantage in CR duration or survival resulted from intensive treatment with DNR and cytarabine following induction of CR. J Clin Oncol 9:2002-2015. © 1991 by American Society of Clinical Oncology.

THE

(CALGB) study" demonstrated a complete remission (CR) rate of 78% in patients treated with DNR, vincristine (VCR), and prednisone, with a median CR duration of 15 months. Approximately one fourth of the responders remained in continuous complete remission (CCR) for 36 months. Induction treatment in the study to be reported was based on that previous CALGB study. Patients who achieved CR were randomly assigned to maintenance treatment as in the earlier study or to intensification treatment with two courses of cytarabine (Ara-C) and DNR, a regimen commonly used in the treatment of acute myelocytic leukemia (AML) in adults. The primary objective of this study (CALGB 8011) was to determine whether this early intensification increased the duration of CR and survival in adults with ALL. To identify significant subgroups of ALL in adults with regard to response frequency and remission duration, ancillary prospective studies were also undertaken of morphologic, cytochemical, cytogenetic, and immunologic markers.

DEVELOPMENT of increasingly effec-

tive treatment for acute lymphocytic leukemia (ALL) in children preceded and continues to

surpass improvement in the treatment of this disease in adults. Major improvement in response rates in adults, unlike children, resulted from the addition of daunorubicin (DNR) to the induction

regimen." A prior Cancer and Leukemia Group B

From Columbia University College of Physicians and Surgeons, New York; Mount Sinai Hospital,New York; New York Hospital-Cornell Medical Center, New York; Roswell Park CancerInstitute, Buffalo, State Universityof New York Health Science Center at Syracuse, Syracuse, NY: University of Chtcago, Chicago, IL; University of Maryland Cancer Center, Baltimore, MD; University of California at San Diego, San Diego, CA: and QualityAssurance Review Center, Providence, RI. Submitted October 2, 1990, acceptedMay 16, 1991. Address repnnt requests to CharlesA. Schiffer, MD, University of Maryland CancerCenter, 22 South Green St, Baltimore, MD 21201. © 1991 by Amencan Society of Clinical Oncology. 0732-183X/91/0911-0009$3.00/0

2002

Journal of Clinical Oncology, Vol 9, No 11 (November), 1991: pp 2002-2015

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2003

INTENSIVE TREATMENT OF ALL IN REMISSION METHODS AND PATIENTS Study Eligibility Patients eligible for the study were individuals over the age of 15 years who had newly diagnosed, previously untreated ALL. All were registered with the CALGB office at the time of entry on study and received identical induction therapy. Subsequent randomization to one of two maintenance arms occurred after the patients had achieved marrow remission and were ready to start intensification. This was accomplished by a closed envelope method, following stratification according to age (15 to 29 years v 30 years and older). Informed consent was obtained from all patients before therapy. The study was approved by the institutional review board at each of the participating institutions.

Treatment Induction phase. The induction treatment was identical to that previously used by CALGB 6 with one exception: the use of early intrathecal (IT) methotrexate (MTX). All patients received VCR 2 mg intravenously (IV) on days 1, 8, and 15 and prednisone 40 mg/m 2 /d, given orally for 3 weeks, which was then tapered over the next 8 days. DNR was given at a dose of 45 mg/m 2 /d IV on the first 3 days. A lower dose of DNR (30 mg/m2i/d) was given to patients aged 60 years and older. Between days 22 and 31, patients received 2 asparaginase, 500 IU/kg/d IV. Initially, IT MTX 12 mg/m (maximum dose, 15 mg) was given on days 1, 8, and 15. It became apparent after 3 months of study accrual that three doses of MTX in this setting induced intolerable toxicity. In February 1981, the protocol was modified, stipulating only one dose of IT MTX on day 1 accompanied by an equal dose of leucovorin given IV at 24 and 48 hours. The first 25 patients treated (before the February 1981 addendum) are excluded from this analysis, as they were treated differently from the remaining cohort. Randomized intensificatton phase. At the point of marrow recovery in those patients who achieved an M1 remission marrow (defined as < 5% blasts in a normocellular specimen containing > 15% nucleated RBCs, > 25% granulocytic cells, and active megakaryopoiesis)-but no earlier than week 6-responders were randomly assigned to receive one of two premaintenance regimens. Regimen 1 consisted of three courses of mercaptopurine (6-MP) and MTX given at 2-week intervals, each course comprising 5 successive days of oral MTX 7.5 mg/m2/d and oral 6-MP 200 mg/m 2/d. This regimen had been used in two prior CALGB studies.'" Regimen 2 was intensive, comprising two courses of Ara-C and DNR. The first course consisted of 7 days of continuous infusion Ara-C 100 mg/m 2 /d accompanied by DNR 45 mg/m /d for the first 3 days (30 mg/m2/d for patients over the age of 60 years). The second, shorter intensification course consisted of 5 days of Ara-C and 2 days of DNR at the same doses. Maintenance phase. Cyclic chemotherapy, started on day 85, was the same for all patients, with each cycle lasting 8 weeks. Each maintenance cycle consisted of a 2-week course of VCR and prednisone followed by three courses of MTX and 6-MP. The VCR dose was 2 mg on days 1 and 8 of each cycle, with prednisone 40 mg/m/id given orally for 14

days and then tapered over the next 8 days. A 5-day course of oral MTX and 6-MP started on the day that prednisone tapering began. There were 9-day, drug-free rest intervals, between courses of 6-MP and MTX. The second total cycle was delayed 1 week to allow the start of CNS prophylaxis. The first course of 6-MP and MTX after cranial radiation comprised the usual 5 days of 6-MP but only 4 days of oral MTX, as it was accompanied by IT MTX. Maintenance treatment was continued until relapse or for 3 years, whichever came first, unless it had to be stopped because of excessive toxicity. CNS Prophylaxis and Therapy A lumbar puncture was scheduled just before or on the day that induction chemotherapy started, with a single dose of IT MTX at the start of induction therapy, followed by an equal dose of leucovorin given IV at 24 and 48 hours. Both IT MTX and cranial radiation were stipulated for all patients achieving an M1 marrow who remained in remission and on study at day 141. This allowed completion of intensification by day 84 and one complete 8-week course of maintenance before CNS prophylaxis. External-beam radiation to the entire brain and the spinal cord above the second cervical vertebra was stipulated. A total midline dose of 24 Gy was planned, with a daily increment of 2 Gy midline and a total elapsed treatment time of 16 days. IT MTX 12 mg/m 2 per dose (15 mg maximum) weekly for 5 weeks was started on the same day as cranial radiation; it was not accompanied by leucovorin. Those patients who did not tolerate IT MTX received IT Ara-C 30 mg/m2 (not to exceed 50 mg total per dose). Patients with evidence of CNS leukemia were to remain on the study. CNS leukemia at the time of diagnosis of ALL was treated immediately with IT MTX, given twice weekly until the CSF cleared, once weekly thereafter for 4 weeks, and once monthly thereafter for a total of 1 year of IT treatment. The patients who responded to IT MTX received cranial radiation, starting 1 week after completion of asparaginase. Subsequent prophylactic CNS treatment (radiation and IT MTX) was not given to such patients. Other patients who developed CNS leukemia without hematologic relapse during the course of systemic treatment received early therapeutic IT MTX and cranial radiation therapy (RT) while remaining on study. Follow-up information regarding CNS disease and CNS treatment was obtained for a majority of study patients. We were able to observe those who did not respond to induction therapy, those responders who did not enter the randomized phase, and those who relapsed after protocol intensification and maintenance to determine the overall rate of CNS relapse. StatisticalConsideration Associations between pretreatment patient characteristics and response to induction and the assessment of comparability of pretreatment characteristics for the two 7 2 randomized groups were evaluated by the Pearson y test. All tests were two-sided with statistical significance set at .05. For premaintenance comparisons, all patients were analyzed as randomized regardless of treatment received,

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2004

ELLISON ET AL

following the policy of intent to treat. Patients with CR who were randomized but never received the premaintenance treatment because of early relapse or refusal were considered canceled for the randomized phase of study and excluded from the comparative analyses of the results of intensive or standard premaintenance therapy. However, such patients are included in analyses of patient characteristics at time of study entry and the comparison of these characteristics with those of patients reaching CR. For purposes of comparison between the two randomized groups, remission duration was defined from date of randomization to date of relapse or death, or last contact with patient in CCR. Survival from date of randomization was defined as the interval from date of randomization to a premaintenance regimen until death or last patient contact. Survival of the entire group of 277 patients was defined as the time from study entry to death or last patient contact. This analysis included all eligible patients registered with the CALGB office who received any induction therapy and for whom data were submitted. Remission duration and survival distributions were estimated by the method of Kaplan and Meier.' The analyses in CR patients included only those with CR who were randomized and treated with one of the two forms of premaintenance therapy. All treatment and subgroup comparisons were performed by the log-rank test.' Multivariate testing for identification of independent prognostic factors for survival and remission duration was conducted by Cox proportional hazards modeling and was performed on a subset of patients for whom all information on pretreatment characteristics was available."'

Immunologic Studies Centralized immunologic studies were initiated in 1983 after reproducibility of results was demonstrated with before and after samples shipped by express mail and in three separate laboratories." Peripheral blood and bone marrow lymphoblast immunophenotypes were then studied. The group-wide studies were initiated 2.5 years after the treatment study started. The immunologic tests required adequate numbers of analyzable cells (bone marrow specimens and/or high WBC and blast levels). Thus, not all patients entered in the treatment protocol could be studied immunologically. The monoclonal antibodies used to determine phenotypes" were B4/Leu-12 (CD19), B1 (CD20), J-5 (CDIO0), 12, T101 (CD5), Tll/Lyt-3 (CD2), 3A1 (CD7), MCS-2 (CD13), and MY9 (CD33). The criterion for surfacemarker positivity was expression in at least 20% of the leukemic blast population. Expression of terminal deoxynucleotidyl transferase (TdT) was determined by an immunofluorescence method. B-antigen expression was defined as B4 or B1 positivity, T-antigen expression as T101 or Tll reactivity, and myeloid-antigen expression as MCS-2 or MY9 positivity. Patients were classified into three major categories: B lineage, T lineage, and those expressing a myeloid antigen (with or without B or T reactivity). Those with no B, T, or myeloid positivity were designated as unclassified. Those with inadequate marker data were termed not classifiable. There was also a small miscellaneous group comprising patients with other combinations of marker positivity.

Data Quality Assurance As part of the data quality assurance program of CALGB, members of the Data Audit Committee make site visits to all institutions at least once every 3 years. Audit Committee members verify compliance with Federal regulations and protocol requirements, including eligibility, treatment, tumor response, and follow-up in a sample of protocols at each institution. A cohort of 51 (17%) patients entered on this study had such on-site review of their medical records.

PatientEntry Between September 1980 and November 1985, 321 patients entered CALGB protocol 8011. Five patients were withdrawn by the institutions (wrong diagnosis, two patients; error in entry, one; no treatment given, two). Nine additional patients were ineligible because of diagnosis (AML, six patients; blastic chronic myelocytic leukemia [CML], two; lymphoma, one) and were not treated on this protocol. Data on five patients were never received. Therefore, 302 eligible patients for whom data were available were entered; 25 of these were treated before the addendum that modified IT MTX therapy and will not be discussed. Thus, we report on 277 eligible patients.

PretreatmentCharacteristics The median age of the 277 patients was 33 years (range, 15 to 79 years). Fifty-nine percent of the patients entered were male. Few of the patients had bulky disease. Although 48% of the patients had some degree of lymphadenopathy, less than 4% had generalized and/or massively enlarged nodes. The spleen was palpable in 39% but extended 6 cm or more below the costal margin in only 12%. The degree of hepatomegaly was similar, with a palpable liver in 37% but enlargement to 6 cm or more below the costal margin in 12%. Although the presence of a mediastinal mass was reported in 9.5%, there was massive abnormality in only 1.5%. The WBC level at the time of study entry ranged from 3,000/piL to 831,000/RL (median, 12,200/RiL). Platelets also ranged widely, from 3,000/pRL to 499,000/p.L (median, 54,000/p[L). The median pretreatment hemoglobin level was 9.5 g/dL (range, 4.4 g/dL to 16.8 g/dL). Three women were pregnant at the time of diagnosis and underwent therapeutic abortions before treatment. A baby had been born to a fourth 2 weeks earlier. One patient with Down's syndrome entered the study. Six patients had a history of prior malignant disease; three of these had received chemotherapy. Three others had prior hematologic abnormalities, with chemotherapy given to one. Central morphologic review" was performed on specimens from 102 patients (Table 1). Submission of bone marrow aspirate and peripheral blood slides was mandated 2 years after the study opened. More than half (57%) of those reviewed were French-American-British (FAB) L2, less than half (39%) were L1, and a few (4%) had unclassified acute leukemia. All patients had fewer than 3% peroxidase-positive blasts. Immunophenotypes, determined by central CALGB laboratory studies of marrow and/or blood, were available on 153 patients. One half of the 153

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2005

INTENSIVE TREATMENT OF ALL IN REMISSION Table 1. Pretreatment Laboratory Characteristics No. of Patients

FAB classification (central review, 102 patients) L1 40 L2 58 UAL 4 Immunophenotype (central review, 153 patients) B lineage 79 21 T lineage Myeloid only 11 Unclassified 12 Miscellaneous 5 BT lineage 4 B myeloid 12 T myeloid 3 BT myeloid 1 Not classifiable 5

%

39 57 4 52 14 7 8 3 3 8 2 1 3

Abbreviation: UAL, unclassified acute leukemia.

infection in two and cardiopulmonary abnormalities during recovery from pancreatitis in the third.

In addition, intensification treatment was delayed by development of abnormal liver function after asparaginase treatment in two others, and marrow relapse occurred at 55 and 65 days, without administration of any premaintenance treatment. This group of 13 patients included only one patient aged younger than 30 years and four aged 60 or older. The response rate was clearly related to the age of the patient (P < .0001; Table 2). While those under the age of 30 years had a 75% CR rate, there was a 64% CR rate in patients aged 30 years through 59 years, and the CR rate was only 33% in those aged 60 years or older. Patients under the age of 50 years usually completed the induction

phase alive, with only 10 of 59 nonresponders patients in whom the immunophenotype of the leukemic cells was determined had B-lineage phenotypes, 14% had T lineage, and the remaining 34% fell into a number of categories. Most interesting was the finding that 18% had myeloid markers, whether as myeloid only, B myeloid, T myeloid, or (in one case) B/T/myeloid. The blasts of those patients found to have a myeloid marker were not peroxidase-positive and did not have any of the morphologic characteristics of myeloblasts or monoblasts. In subsequent analyses, the 10 patients with nonspecific phenotypes (miscellaneous or not classifiable) will not be included. At study entry, 21 of the 277 patients (7.6%) had CSF abnormalities, suggesting the presence of meningeal leukemia, although only one initially had symptoms of CNS disease and one other developed a seventh-nerve palsy during induction. The presence of a mediastinal mass in 29% of these patients was the only difference from the group as a whole. All of the 21 patients with CNS or CSF abnormalities at the time of diagnosis were treated as stipulated by the protocol; 16 of the 21 achieved M1 marrows. However, due to patient or physician choice, only nine were randomized for the intensification phase and treated accordingly: four on one arm of the study and five on the other. RESULTS Induction Response Rate and Pretreatment Characteristics A CR was achieved by 177 patients (64%) (Table 2). An additional 13 patients achieved M1 marrows but are not included in those categorized as CRs. Three of these 13 patients never attained H1 blood status; one of these relapsed and two died 10 to 26 days after the M1 marrow was noted. Eight of the 13 patients achieved M1, H1 ratings, but five had rapid relapse (13 to 34 days) and three did not relapse but died early (12 to 38 days), with

dying during induction. In contrast, 28 of the 41 nonresponders over the age of 50 years died during the induction period, usually of complications related to pancytopenia. The only factor other than age and, to a lesser extent, pretreatment WBC level that appeared to be of importance was the presence of a myeloid Table 2. Response Rate CR Entry Charadcteristics

Total

No

%

All Age (years) 15-29 30-59 > 60 Hepatomegaly Present Absent Splenomegaly Present Absent WBC/.L < 50,000 > 50,000 Platelets/tL < 50,000 > 50,000 FAB type L1 L2 UAL Immunophenotype* Myeloid+ Myeloid-

277

177

64

P

114 120 43

86 77 14

75 64 33

< .0001

100 175

58 118

58 67

.12

106 168

69 107

65 64

.81

198 79

134 43

68 54

.04

133 144

81 96

61 67

.32

40 56 4

28 35 1

70 63 25

.46

27 116

8 90

36 78

< .0001

*Myeloid antigens present (+) or absent ( ) on leukemic blast cells.

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2006

ELLISON ET AL

marker on the leukemic cells. Patients with a myeloid marker, either alone or associated with a lymphoid marker, achieved only a 30% CR rate, whereas there was a 72% CR rate among those without myeloid markers. Numerous other factors (Table 2) including sex, mediastinal mass, presence or absence of CNS disease, splenomegaly, hepatomegaly, lymphadenopathy, pretreatment platelet count, and the FAB classification showed no relationship to the CR rate. Induction Toxicity and Protocol Compliance The initial treatment with DNR, VCR, and prednisone resulted in marked myelosuppression. Life-threatening toxicity (granulocytes < 500/[L or platelets < 25,000/p*L) occurred in 243 patients (88%), and 24 patients (9%) died during the initial phase of induction. Only four patients had DNR dose modifications of any type, and 7% had VCR dose alterations. Gastrointestinal toxicity, liver function test abnormalities, and renal abnormalities were rare, usually occurring in the presence of hematologic complications and infection. Tumor lysis syndrome was reported in four patients and disseminated intravascular coagulation in three. Asparaginase was scheduled to start on day 22 of induction; it was started within 1 day of the scheduled time in 85% of the 239 patients who received this drug and within 1 week in 94%. Of the 239 patients who received asparaginase, 103 had already attained an M1 marrow. A partial remission (M2 marrow with 6% to 25% blasts in a cellular specimen) was present in 31 patients, 17 of whom subsequently achieved CRs. CR also developed after asparaginase treatment in 32 of 40 patients whose drug course started in the presence of a marrow aspirate depleted of blasts but lacking sufficient regeneration of myeloid and/or erythroid precursors to designate this as a partial remission. An additional 65 patients started asparaginase with persistent leukemia in the marrow; 34 of these patients achieved CRs. While the asparaginase phase was better tolerated than the initial phase of induction therapy, it still caused severe, life-threatening, or fatal toxicity in two thirds of the patients; the most frequent major adverse effects were leukopenia (49%) and liver function abnormalities (21%). Pancreatitis, gastrointestinal toxicity, and hyperglycemia were rare. Anaphylaxis occurred in five patients; three

subsequently received a complete course of Erwinia asparaginase. Seven patients died during asparaginase toxicity: three with sepsis, two with hepatic toxicity, one with infection as well as renal and hepatic toxicity, and one with cardiopulmonary abnormalities. These deaths occurred in five patients over the age of 50 years and in two aged 34 years and 40 years. Because of the degree of toxicity, 35 patients received fewer than the stipulated 10 doses of asparaginase, but 24 of these 35 patients received seven to nine doses. IntensificationPhase Although 177 patients achieved CR, 26 of these were withdrawn from the study and did not enter the intensification phase (Table 3). Physician decision resulted in withdrawal of four of these patients before randomization: one patient (age 31 years) required a gastrectomy for a bleeding ulcer; a second (age 23 years) was removed from study on day 22 for "intensive therapy" not otherwise specified; the third (age 24 years) was withdrawn on day 32 for allogeneic bone marrow transplantation; and the fourth (age 48 years) was removed from study on day 82 because of the degree and duration of toxic effects after asparaginase administration. Physician error led to cancellation of another three patients, as randomization was erroneously not done in two (ages 20 and 23 years) and the dose of asparaginase was only 10% of that stipulated in a third (age 17 years), who was therefore not randomized. Six patients (ages 20, 32, 42, 56, 59, and 71 years) refused randomization, and seven others (ages 25, 26, 29, 33, 34, 48, and 72 years) refused treatment after randomization. The reason was not apparent for withdrawal of six other patients (ages 20, 22, 30, 31, 45, and Table 3. Premaintenance Randomization No of Patients

Achieved M1 marrow Achieved CR Not randomized, off study Physician decision Physician error Patient refused Cause unknown Randomized, not treated Patient refused Randomized, treated Standard therapy Intensive premaintenance

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

190

177 19 4 3 6 6 7 7 151 77 74

2007

INTENSIVE TREATMENT OF ALL IN REMISSION 58) before randomization. Since these 26 patients received no protocol therapy after CR, they are considered as cancelled randomizations and are excluded from analyses of premaintenance effects. Thus, there were 151 patients who accepted randomization and were treated: 77 randomized to the standard therapy and 74 to the more intensive treatment. Two of the latter actually received treatment with the standard maintenance treatment-one in error and the other because of cardiac abnormalities (a low ventricular ejection fraction)--but both are analyzed in the group to which they were randomized. The pretreatment clinical and laboratory characteristics of the two treatment groups did not differ significantly (Table 4). For no obvious reason, the groups differed, however, in the time at which they were randomized to start and actually started on premaintenance treatment, with therapy delayed at least 2 weeks in 10% of the standard group and 32% of the intensively treated patients. This difference was not clearly related to the differences noted in the degree of asparaginase toxicity in the two groups, since there was only a trend towards an increased percentage of patients with severe or life-threatening toxicity in those randomized to receive intensive premaintenance (70% v 59%). There were marked differences in the toxic effects of the two premaintenance regimens (Table 5). The more intensive regimen produced major myelosuppression in all patients and lifeTable 4. Intensification Phase: Pretreatment Characteristics of the Two Treatment Groups Standard (n = 77)

Sex Male Female Age (years) 15-29 30-59 > 60 CNS disease Hepatomegaly Splenomegaly Lymphadenopathy WBC/[L < 50,000 > 50,000 Platelets/piL < 50,000 > 50,000

Intensive (n = 74)

No

%

No

%

P

46 31

60 40

42 32

57 43

.64

38 34 5 4 21 28 32

49 44 6 5 27 36 41

38 29 7 5 30 35 40

51 39 9 7 41 48 54

.75 .48 .09 .19 .14

59 18

77 23

57 17

77 23

.92

37 40

48 52

35 39

47 53

.87

Table 5. Intensification Phase Toxicity None/ Mild

Standard regimen WBC and/or platelet levels Infection Other Overall maximum toxicity Intensive regimen WBC and/or platelet levels Infection Other Overall maximum toxicity

LifeModerate Severe Threatening Fatal

51 85 57

23 14 27

19 1 16

7 0 0

0 0 0

35

32

26

7

0

0 17 60

0 25 21

0 32 13

97 21 3

3 4 3

0

0

0

96

4

NOTE. Data are percentages.

threatening or fatal infection in 25%. Severe gastrointestinal toxicity also occurred in 15% and hepatic function abnormality in 3%. Three patients (4%) died during this phase of the protocol, all from infection while neutropenic. Eight patients were treated with only one intensive course because of the degree and/or duration of toxicity. The protocol called for a 28-day interval from the first day of course 1 to the first day of course 2 of the intensive premaintenance treatment. The degree of toxicity often delayed this, however, and the interval actually ranged from 22 to 57 days (median, 34 days). The standard regimen was far less toxic. Although one third of the patients had severe or life-threatening toxicity (hematologic or gastrointestinal), this was life-threatening in only 7%; most of the toxic effects were rated as severe, and there were no fatalities. The difference in the degree of toxicity in the two arms resulted in further delay in starting maintenance therapy. In those who received the intensive regimen, 63% had delays longer than 2 weeks, compared with 46% on standard therapy. Maintenance Phase A total of 137 patients entered the maintenance phase; 72 had previously received the standard premaintenance chemotherapy, and 65 had been treated with the intensive regimen. Maintenance chemotherapy produced some toxicity in the majority of patients, with the side effects somewhat greater in those who had previously received intensive chemotherapy. These pa-

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2008

ELLISON ET AL

tients not only had more mucositis but twice as many had some degree of early hematologic toxicity. However, life-threatening toxicity was rare in both groups. Infections were infrequent, and the neurologic manifestations were almost completely limited to paresthesias. Due to toxic effects, modifications were made in dose or schedule of 6-MP and/or MTX in two thirds of the patients.

1

0.75 >l-m

0.50

CL 0 0.

0.25

CNS Prophylaxis Table 6 shows the flow of the patients through the various stages of the study, indicating that 112 patients received the stipulated CNS prophylaxis. Sixty-one of these had been treated with standard premaintenance and 51 with the intensive chemotherapy. The median time for starting CNS prophylaxis in those who received standard treatment was 151 days after entry on study, only 10 days later than planned, although 21% of the patients were at least 3 weeks late in starting. As might be expected, the delay was far more striking in the group that had received intensive premaintenance. The median time at which CNS prophylaxis was started in the latter was 165 days after study entry; one third of the patients started more than 6 weeks late. Remission DurationFrom Randomization The median follow-up for response duration is 6.1 years. Remission duration (Fig 1) is identical for the two treatment groups (P = .92), with an overall median of 1.75 years. CR lasted longer than 3 years from randomization in 40% of the patients randomized and treated on the two arms of the study. No relapse has yet occurred later than 60 months. Twenty-nine percent of those who entered the randomized phase remain in CCR. The 44 patients still in CCR are evenly divided Table 6. CNS Prophylactic Therapy

Total no. of patients entering intensification Reason for omitting CNS prophylaxis CNS disease pretreatment and treated CNS disease diagnosed and treated During induction During intensification During first week of CNS prophylaxis Died in CR before prophylaxis BM relapse, taken off study No. of patients receiving prophylaxis

Standard

Intensive

77

74

4

5

1 5 0 0 6 61

2 6 2 2 6 51

0 0

1

2

3

4

5

6

7

8

9

10

YEARS Fig 1. CR duration of patients randomized to two premaintenance regimens: standard (-; censored, 22; failed, 55; total, 77; median, 1.71 years) or intensive (----; censored, 22;

failed, 52; total, 74; median 1.78 years). P = .92.

between treatment with standard premaintenance and intensive therapy. The longest CCR so far is 117 months, with 19 patients in CCR longer than 7 years. The CR duration was significantly longer (P = .0118) for those under the age of 30 years (Fig 2), with a median duration of 2.84 years compared with 1.12 years in those aged 30 to 59 years, and 1.14 years in those 60 years or older. The majority of those remaining in remission are in the youngest age group. Just over half (53%) of the youngest group of patients (ages 15 through 19 years) who received randomized protocol premaintenance remain in CR at last report, compared with 25% of those aged 20 through 39 years and 18% of those aged 40 years and over. There was no difference in median CR duration between those aged 20 through 39 years and those aged 40 through 59 years. Only one patient over 60 years of age (8%) remains in CR. Prognostic factors identified other than age include the WBC count at the time of study entry (P < .0001) and the pretreatment platelet count (P = .008). Those patients who achieved a CR and entered the intensification phase after attainment of an M1 marrow within 30 days from study entry had better results than did those whose M1 marrow was reached later than 30 days (P = .03; Table 7). This relationship was noted in both the standard and intensive premaintenance arms. Survival of Patients Who Received Premaintenance Survival from time of randomization for the patients treated on the two premaintenance arms

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2009

INTENSIVE TREATMENT OF ALL IN REMISSION

0.75 I-J

m 0.50 m 0

u--i

...

. ..

......

..

7

8

0.

0.25 I -1

0 2

3

4

5

6

9

10

YEARS Fig 2. Effect of age on CR duration: 15 to 29 years (-; censored, 29; failed, 47; total, 76; median, 2.84 years); 30 to 59 years (----; censored, 14; failed, 49; total, 63; median, 1.12 years); 60 years or older (.....; censored, 1; failed, 11; total, 12; median, 1.14 years). P = .0118.

was the same, with medians of 2.78 and 2.74 years (P = .77; Fig 3). Subgroups by age, WBC count, and platelet count had significantly different prognoses (P = .004, P < .0001, and P = .008, respectively), with higher WBC level and lower platelet counts having a poorer prognosis. Multivariate analysis by Cox regression indicated that age, pretreatment WBC level, and platelet count were each significant, independent prognostic factors. OverallSurvival The median follow-up for survival is 5.6 years. The median survival for all 277 eligible patients (Fig 4) was 1.32 years. Multivariate analysis of overall survival by Cox regression was performed on a subset of 270 patients who had complete data with regard to pretreatment characteristics, with the exception of FAB and immunophenotype, for which data are known for a smaller number of patients. This analysis of continuous variables showed that increasing age (P < .0001; Fig 5) and higher WBC count (P < .0001) were significant independent factors predictive of poorer survival, as was the presence of a myeloid marker (P = .001) in a subgroup of 143 patients. However, survival from time of premaintenance for the few with myeloid markers who achieved CRs does not differ from survival for those without such a marker. CNS Relapse To date, 68 (25%) of the 277 patients who entered the study are known to have had CNS

abnormality at some time, including 21 at the time of entry on study. Fourteen of these 21 achieved CRs; 10 of the 14 entered the randomized phase, and three of the 10 remain in CCR without any further evidence of CNS abnormality. Subsequently, only one of the 21 patients with CNS abnormality pretreatment is known to have developed CNS disease, which appeared with concurrent bone marrow (BM) relapse 1 month after randomization. This patient was then removed from the study and survived another 9 months. Twenty-nine patients developed CNS disease before receiving, CNS prophylaxis. Five of the 29 developed CNS disease during induction. Two of the five did not attain CR; the three who achieved CRs entered the randomized intensification phase but had BM relapse within 6 months and subsequently died. Of the 29 patients who were not randomized for premaintenance (mostly because of lack of CR) and who did not receive CNS prophylaxis at any time, 10 developed CNS leukemia after being taken off study; all 10 have died. Of the 151 patients randomized for premaintenance, 11 developed CNS leukemia before CNS prophylaxis. Although 10 of these 11 then remained on protocol treatment (as stipulated), all have since had BM relapse, and all have died. One additional patient was removed from study because of BM relapse before the time CNS prophylaxis was due and subsequently developed CNS disease. Two others were noted to have CNS disease within a few days after the start of CNS prophylaxis but remained on study; both eventually had BM relapse and died. Thus, none of the 29 who developed CNS disease before receiving any CNS prophylaxis is still alive. Despite the administration of CNS prophylaxis per protocol to 112 patients, meningeal leukemia subsequently developed in 18: in 11 patients while they were still on protocol treatment or follow-up Table 7. Duration of CR Versus Time to M1 Marrow Time to Achieve MI Marrow (days) < 30

> 30

Median time to M1 marrow, days (range) 22 (18-29) 43 (31-126) No. of patients 82 69 Median duration of CR, days (range) 921 (50-3,040+) 525 (15-3,043+) CR > 3 years (%) 48 30 P = .03

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2010

ELLISON ET AL 1I

0.75 >-J

m

0.50

m 0 0.

0.25

0

1

2

3

4

5

6

7

8

9

10

YEARS Fig 3. Survival of responders after randomization to two premaintenance regimens: standard (-, alive, 29; dead, 48; total, 77; median, 2.78 years) or intensive (----; alive, 28; dead, 46; total, 74; median, 2.74 years). P = .77.

in CCR, and in seven after they had exited the protocol because of BM relapse. Four of the 11 with CNS relapse while on protocol had concurrent BM relapse and were appropriately removed from the study at that time. Despite lack of BM relapse, three others were also removed from study when CNS disease was diagnosed: one was lost to follow-up 5 months later, one had BM relapse 3 months later and has since died, and the third is still alive 4 months after removal from study. Thus, CNS leukemia occurred despite CNS prophylaxis in 21% (13 of 61) of those who had received the standard premaintenance but in only 10% (five of 51) of those who received intensive therapy (P = .054). CNS disease developed 5 to 42 months after the start of CNS prophylaxis on protocol. CNS relapse occurred concurrent with BM relapse in seven of these 18 relapses. Only one patient is known to be alive after CNS relapse but is not included among those still in CCR.

still in BM remission: one after induction resulted in CR and two after standard premaintenance therapy. All three died 3 to 8 months after being taken off study. Ten patients who had either rapid relapse of an M1 marrow (two patients), initial partial marrow response (five), or no response (three) had BMT after being taken off study. Two of this group remain alive at 55 and 68 months. One patient with an M1 marrow but CNS relapse at 24 months underwent autologous BMT and was known to be alive 5 months later. One patient with an M1 marrow but testicular relapse at 41 months also underwent autologous BMT; he died 8 months later. Eleven patients with BM relapse 10 to 51 months after study entry received marrow transplants: four autologous and seven allogeneic. None remains alive. Late Toxic Effects Aseptic necrosis of the femoral head and/or the tibial plateau occurred in five patients: four males and one female. All of the males had been treated with intensive premaintenance chemotherapy; the female had received standard treatment. One 24-year-old female developed cataracts. One man developed acute renal failure requiring dialysis, and another developed status epilepticus due to Aspergillus brain abscesses. Fungal infection in another patient necessitated a partial nasal septectomy. These patients all remain in remission. PregnancyAfter Completion of Protocol Treatment There have been at least eight pregnancies in women treated on this protocol: one patient had

Other Non-BM Relapses Testicular relapses occurred 13 to 43 months after diagnosis of leukemia in six (ages 19 to 44 years) of the 88 men treated in the intensification phase (three treated on each regimen). Testicular relapse was concurrent with BM relapse in two of the six. One patient developed a neck mass, biopsy-proven to be leukemic nodes, without other evidence of relapse. BM Transplantation Three of the patients were removed from the study to undergo BM transplantation (BMT) while

0.75 l-

S0.50 m

0.25

0

0

1

2

3

4

5

6

7

8

9

10

YEARS Fig 4. Survival of all patients from time of study entry: alive, 71; dead, 206; total, 277; median, 1.32 years.

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2011

INTENSIVE TREATMENT OF ALL IN REMISSION I

remission duration or the proportion of long-term remissions. A recent review analyzed treatment results and compared the numerous reported studies in chil-

0.75 >-I m

dren and adults.3 6 There is general acceptance of

the prognostic importance of several pretreatment

0.50

m

characteristics in ALL of children, 37"38 including

o

a-

0.25

o 0

1

2

3

4

5

6

7

8

9

10

YEARS Fig 5. Effect of age on survival of all patients entered in study: 15 to 29 ycears (-; alive, 43; dead, 71; total, 114; median, 2.64 years ); 30 to 59 years (----; alive, 27; dead, 93; total, 120; median 1.17 years); and 60 years or older (.....; alive, 1; dead, 42; total, 43; median, 0.14 years). P < .0001.

an abortion, an d seven other pregnancies went to term successfullyly. DISCUSSION The use of multidrug, multicycle chemotherapy for the treatm ent of ALL in adults and the continuation of treatment for at least 2 or 3 years result from the major improvements in both remission duration ai nd ultimate survival that have been obtained in childldren using such treatment strate6 gies.14-20 Most r eported studies in adult ALL ,21-34 were small and /or were not designed as randomized comparatii ve trials. Although high response rates were ustally noted, there were marked variations in r esponse duration, with medians ranging from 20 to 57 months. However, comparison of these re ports is difficult because of wide variations in paitient age and criteria for eligibility, as well as excl ision of Ph+ patients from some studies. The study re ported here is the only prospectively randomiz ed investigation of the effects of intensive postr emission chemotherapy in adults. The combinationn of Ara-C and DNR was chosen because of del monstrated effectiveness of each agent as induct ion treatment in ALL, when used 35 in combination' 1-6or alone. The choice of doses and schedule was based on the broad experience of CALGB in th e use of this combination, albeit in AML. The com bination of Ara-C and DNR given twice as intensification resulted in major toxic effects in most patients but did not affect either

age, lymphoblast morphology, leukocyte count, presence or absence of bulky disease, immunophenotype, cytogenetics, and the time to achieve CR. Reports on prognostic factors in ALL of adults are more contradictory. The inconsistency of reported prognostic factors with regard to response rate, remission duration, and survival is striking. 21-2331-34 Generally speaking, however, the CR rate was related inversely to patient age, although the cut-off for major change ranged from ages 25 to 40

years. The pretreatment WBC count is reported by some to affect the CR rate but relates more 30 '31but not consistently to CR duration. Most 22-24, all investigators have reported that those patients who reach CR rapidly have longer CRs. Beyond this, there is even greater disagreement; T-cell markers, particularly when associated with a mediastinal mass, have been reported as adverse,31' but not by all.19,22,29 There are also discordant reports on the effects of FAB category, lactate dehydrogenase level, sex, degree of BM involvement, percent of circulating blasts, hepatomegaly, splenomegaly, etc. Some reports related prognosis to the total number of adverse factors present.31'34 This study was designed to test the effects in adults with ALL of two intensive courses of Ara-C and DNR after CR induction by DNR, VCR, prednisone, and asparaginase. Comparison of patients who received two courses of intensive chemotherapy with those who were treated less agressively indicated no effects on either the duration of CR, the proportion still remaining in CCR, or survival. Twenty-nine percent of all who entered the randomized phase remain in remission 3.5 to 9.6 years after randomization: 37% of those younger than 30 years and 23% of those 30 or older. The median duration of CR for all was 22 months: 36 months for those under age 30 years and 13 months for all others. The pretreatment WBC and platelet levels and time to achieve CR also affected remission duration. The presence of a myeloid marker was prognostically significant for survival in the entire group of 277 patients, as were age, WBC count, and platelet count.

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2012

ELLISON ET AL

It is difficult to compare the results of different studies in ALL, as comparisons involve not only the treatment regimens but the populations of patients treated (Table 8). While it is possible that some forms of treatment supersede the prognostic features thought to be important, it is difficult to prove this without direct comparison of two treatments in a randomized study. The median age of the patients in the current study was higher than that in any other series reported to date, with the exception of the group with a median age of 37 years described by Radford et al.3 While the distribution of some other pretreatment characteristics was similar, detailed information concerning WBC and platelet levels and immunophenotyping is not available for these patients. The duration of CR, the overall survival, and the current survival plateau are impressive. The induction regimen used was similar to the CALGB program. It is not clear which component may have contributed to the overall results in the study reported by Radford et al or whether any apparent difference is related to the small number of patients reported. CNS prophylaxis was early (day 32 or day 45) in the series reported by Radford et al. The importance of chromosome analysis in ALL has been reported.3' The effects on median duration of CR and median survival were striking

in children, with karyotype found to be an independent prognostic factor. CALGB is undertaking prospective cytogenetic analysis in newly diagnosed ALL in adults entered on successive CALGB treatment protocols. A recent report" presents cytogenetic information obtained to date. The findings on 80 patients (only a few of whom were in the drug trial reported here) indicate that 23% had a Ph chromosome; presence of this abnormality was associated with older age, expression of B-lineage-associated antigens, inferior CR rate, and few long-term survivors. Treatment studies done by others frequently and systematically delete patients shown to have a Ph chromosome, rendering comparison even more difficult. Prospective CALGB studies of the immunophenotype of ALL lymphoblasts in adults were initiated after the start of this study, with appropriate samples obtained in 55% of patients (Table 1).11'12,1 The number of patients possessing myeloid antigens, usually in conjunction with B-cell or T-cell antigens,"2 was unexpectedly high, and these patients had a poorer prognosis. While essentially the same method of CNS prophylaxis was used in this study as in two prior CALGB trials,"' the timing was different, with treatment given far later in this study. A total of 21 patients (7.6%) had CNS abnormalities on entry to the study; 19 patients had CSF abnormality

Table 8. Comparison of Results Protocol (reference no)

No. patients Age (years) Range Median % CR CR duration Median (months) Plateau (%) Last relapse (months) Overall survival Median (months) Plateau (%)

L-2, L-10, 10M/

SWOG

HOP-L (33)

Bay (28)

01-81 (30,31)

02-84 (30)

L-17, 17M (34)

L-10M (32)

8011

59

81

368

380

199

168

277

15-74 37 75

16-48 24 94

15-65 25 74

15-65 80

> 15 * 82

18-85 28 68

15-79 33 69

50 43 83

51 NR 51

25 36 63

NR 54 Too soon

28 37 66

23 30 58

21 28 60

28 35

52 NR

28 35

NR 53

31 36

18 28

16 20

Abbreviations: HOP-L, doxorubicin, VCR, prednisone, and asparaginase; Bay, California Group Study using DNR, VCR, prednisone, asparaginase, teniposide, and cytarabine, 01-81 and 02-84, German Cooperative Group studies of cyclic combinations of DNR, VCR, prednisone, asporaginose, cyclophosphamide, cytarabine and 6-MP; L-2, L-10, 10M/L-17, 17M, a series of multidrug protocols undertaken at Memorial Sloan-Kettering Cancer Center between 1969 and 1985; SWOG, Southwest Oncology Group; NR, not reached at time of report. 23 *Median age not given, but previous report on 149 of the patients indicated a median age of 25 years.

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2013

INTENSIVE TREATMENT OF ALL IN REMISSION without symptoms before treatment and two had CSF abnormality with symptoms before or early in treatment. This was similar to the 10% incidence reported by Gottlieb et al' and the 6.5% noted by Hoelzer et al,30 but was higher than the 2.8% in patients analyzed by Kantarjian et a142 for CNS leukemia risk factors. An additional 29 patients (10.5%) developed CNS disease before starting CNS prophylaxis. As 14 of these developed CNS disease after randomization but before CNS prophylaxis, it is conceivable that this relapse might not have occurred had prophylaxis started earlier. Omura et a43,44 reported that CNS prophylaxis given after induction and consolidation treatment (ie, at 7 to 9 weeks) in adults with ALL resulted in significantly more CNS leukemia in patients who did not receive prophylaxis. There was no effect on hematologic remission duration or survival. In our study, 18 of 112 patients (16%) developed CNS leukemia 5 to 42 months after receiving CNS prophylaxis. Thus, there were 68 patients (25%) with CNS abnormality at some point. There was a trend towards a lower rate in those who received the intensive regimen. This suggests possible benefit due to Ara-C entry into the CSF during the intensification phase. The real overall rate of CNS leukemia may actually be higher, as 41 of the 44 patients still being observed on protocol in CCR (three having had initial CNS abnormality) remain at risk for development of meningeal leukemia. It appears essential that some type of CNS prophylaxis be given earlier in the course of intensification or other cyclic treatment. The current CALGB study in adults with ALL (which contains more intensive cyclic induction and intensification therapy) includes administration of IT MTX at 4 and 8 weeks, followed by cranial RT and a 5-day course of IT MTX at 12 weeks. Gottlieb et al6 reported 4% CNS disease in the CR group but did not provide information concerning development of CNS leukemia after removal

from study of either the responders or nonresponders. Hoelzer et a130 noted a 14% CNS relapse rate (with 11% as the primary relapse site) despite early use of RT and IT MTX. Kantarjian et a142 indicated 16% CNS disease after treatment with an intensive, multicyclic chemotherapy regimen (VCR, Adriamycin [doxorubicin; Adria Laboratories, Columbus, OH], dexamethasone, cyclophosphamide, MTX, Ara-C, DNR, 6-MP, prednisone, etoposide, and carmustine) that included agents with potentially good penetration of the bloodbrain barrier or drugs given in high-doses; patients did not receive RT or IT chemotherapy. He also reported that less intensive treatment regimens resulted in 27% CNS relapses, for an overall rate of 20%. In conclusion, the use of a 2-month intensification course of Ara-C and DNR in adults with ALL did not improve remission duration, the percentage of patients remaining in first CR, or survival when compared with patients randomly assigned to receive standard maintenance. Comparison of our results with those of other studies shows similarities when age differences and other prognostic factors are considered. The failure of this particular combination and schedule of drugs does not mean that other trials of intensification and/or maintenance therapy are not essential. However, it does indicate that when an aggressive treatment regimen appears superior to earlier strategies, meaningful randomized comparison of the new and the older treatments is important, especially when the standard schedule is better tolerated. It is also important to stratify treatment groups by

prognostic characteristics, particularly age and immunophenotype, and to continue to study immunologic and cytogenetic features. ACKNOWLEDGMENT The authors and CALGB acknowledge with gratitude the major role played by the late Dr Arlan J. Gottlieb in the formulation and implementation of this study.

APPENDIX The following are the participatinginstitutions (and the National Cancer Institute grants supporting this study): Bowman Gray School of Medicine, Winston-Salem, NC (CA 03927); Central Massachusets Oncology Group, Worcester, MA (CA 37135); Columbia University College of Physicians and Surgeons, New York, NY (CA 12011); Dartmouth-Hitchcock Medical Center, Hanover, NH (CA 04326); Finsen Institute, Copenhagen, Denmark; CALGB Central Office, Lebanon, NH (CA 31946); CALGB Immunology and Genetics Committee, Buffalo, NY (CA 37027); Harvard School of Public Health, Boston, MA (CA 33601); Long Island Jewish Medical Center, New Hyde Park, NY (CA 11028); Massachusets General Hospital, Boston, MA (CA 12449); McGill Medical Center, Montreal, Quebec, Canada (CA 31809); Mount Sinai Hospital, New York, NY (CA 04457);

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2014

ELLISON ET AL APPENDIX (Cont'd)

New York Hospital-Cornell Medical Center, New York, NY (CA 07968); Quality Assurance Review Center, Providence, RI (CA 29511); Rhode Island Hospital, Providence, RI (CA 08025); Roswell Park Cancer Institute, Buffalo, NY (CA 02599); State University of New York Health Sciences Center-Downstate, Brooklyn, NY (CA 25119); State University of New York Health Science Center at Syracuse, Syracuse, NY (CA 21060); University of California at San Diego, San Diego, CA (CA 11789); University of Chicago, Chicago, IL (CA 41287); University of Maryland Cancer Center, Baltimore, MD (CA 31983); University of Minnesota, Minneapolis, MN (CA 16450); University of Missouri, Columbia, MO (CA 12046); Walter Reed Army Medical Center, Washington, DC (CA 26806); West Virginia University Medical Center, Morgantown, WV (CA 28562); and Wilmington Medical Center, Wilmington, DE (CA 37041).

REFERENCES 1. Henderson ES, Scharlau C, Cooper MR, et al: Combination chemotherapy and radiotherapy for acute lymphocytic leukemia in adults: Results of CALGB protocol 7113. Leuk Res 3:395-404, 1979 2. Willemze R, Hillen H, Hartgrink-Groeneveld CA, et al: Treatment of acute lymphoblastic leukemia in adolescents and adults: A retrospective study of 41 patients (1970-1973). Blood 46:823-834, 1975 3. Gee TS, Haghbin M, Dowling MD, et al: Acute lymphoblastic leukemia in adults and children. Differences in response with similar therapeutic regimens. Cancer 37:1256-1264, 1976 4. Shaw RT, Raab SO: Adriamycin in combination chemotherapy of adult ALL: A Southwest Oncology Group study. Med Pediatr Oncol 3:261-266, 1977 5. Lister TA, Whitehouse JMA, Beard MEJ, et al: Combination chemotherapy for acute lymphoblastic leukemia in adults. Br Med J 1:199-203, 1978 6. Gottlieb AJ, Weinberg V, Ellison RR, et al: Efficacy of daunorubicin in the therapy of adult acute lymphocytic leukemia: A prospective randomized trial by Cancer and Leukemia Group B. Blood 64:267-274, 1984 7. Snedecor GW, Cochran WG: Statistical Methods (7th ed). Ames, IA, Iowa State University Press, 1980 8. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958 9. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50:163-170, 1966 10. Cox DR: Regression model and lifetables. J R Stat Soc [B] 34:187-220, 1972 11. Sobol RE, Mick R, LeBien TW, et al: The reproducibility of lymphocytic leukemia phenotype determinations: Evaluation of monoclonal antibody and conventional hematopoietic markers. Leuk Res 10:481-485, 1986 12. Sobol RE, Mick R, Royston I, et al: Clinical importance of myeloid antigen expression in adult acute lymphoblastic leukemia. N Engl J Med 316:1111-1117, 1987 13. Davey FR, Mick R, Nelson DA, et al: Morphologic and cytochemical characterization of adult lymphoid leukemias which express myeloid antigen. Leukemia 2:420-426, 1988 14. Riehm H, Gadner H, Henze G, et al: The Berlin childhood acute lymphoblastic leukemia therapy study, 1970-1976. Am J Ped Hematol Oncol 2:299-306, 1980 15. Sallan SE, Hitchock-Bryan S, Gelber R, et al: Influence of intensive asparaginase in the treatment of childhood

non-T-cell acute lymphoblastic leukemia. Cancer Res 43: 5601-5607, 1983 16. Clavell LA, Gelber RD, Cohen HJ, et al: Four-agent induction and intensive asparaginase therapy for treatment of childhood acute lymphoblastic leukemia. N Engl J Med 315:657-663, 1986 17. Steinherz PG, Gaynon P, Miller DR, et al: Improved disease-free survival of children with acute lymphoblastic leukemia at high risk for early relapse with the New York regimen-A new intensive therapy protocol: A report from the Childrens Cancer Study group. J Clin Oncol 4:744-752, 1986 18. Freeman AI: "High-risk" acute lymphocytic leukemia. J Clin Oncol 5:997-998, 1987 (editorial) 19. Koizumi S, Fujimoto T, Takeda T, et al: Comparison of intermittent or continuous methotrexate plus 6-mercaptopurine in regimens for standard-risk acute lymphoblastic leukemia in childhood (JCCLSG-S811). Cancer 61:12921300, 1988 20. Miller DR, Leiken SL, Albo VC, et al: Three versus five years of maintenance therapy are equivalent in childhood acute lymphoblastic leukemia: A report from the Childrens Cancer Study group. J Clin Oncol 7:316-375, 1989 21. Baccarani M, Corbelli G, Amadori S, et al: Adolescent and adult acute lymphoblastic leukemia: Prognostic features and outcome of therapy. A study of 293 patients. Blood 60:677-684, 1982 22. Schauer P, Arlin ZA, Mertelsmann R, et al: Treatment of acute lymphoblastic leukemia in adults: Results of the L-10 and L-10M protocols. J Clin Oncol 1:462-470, 1983 23. Hoelzer D, Thiel E, Lffiler H, et al: Intensified therapy in acute lymphoblastic and acute undifferentiated leukemia in adults. Blood 64:38-47, 1984 24. Clarkson B, Ellis S, Little C, et al: Acute lymphoblastic leukemia in adults. Semin Oncol 12:160-179, 1985 25. Sinchez-Fayos J, Outeirifio J, Villalobos E, et al: Acute lymphoblastic leukaemia in adults: Results of a 'total-therapy' programme in 47 patients over 15 years old. Br J Haematol 59:689-696, 1985 26. Marcus RE, Catovsky D, Johnson SA, et al: Adult acute lymphoblastic leukemia: A study of prognostic features and response to treatment over a ten year period. Br J Cancer 53:175-180, 1986 27. Barnett MJ, Greaves MF, Amess JAL, et al: Treatment of acute lymphoblastic leukaemia in adults. Br J Haematol 64:455-469, 1986

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.

2015

INTENSIVE TREATMENT OF ALL IN REMISSION 28. Linker CA, Levitt LJ, O'Donnell M, et al: Improved results of treatment of adult acute lymphoblastic leukemia. Blood 69:1242-1248, 1987 29. Peters WG, Willemze R, Colly LP: Intermediate and high-dose cytosine arabinoside-containing regimens for induction and consolidation therapy for patients with acute lymphoblastic leukemia and lymphoblastic non-Hodgkin's lymphoma: The Leyden experience and review of the literature. Semin Oncol 14:86-91, 1987 30. Hoelzer D, Thiel E, Lbffler H, et al: Teniposide (VM-26) and cytosine arabinoside as consolidation therapy in adult high-risk patients with acute lymphoblastic leukemia. Semin Oncol 14:92-97, 1987 31. Hoelzer D, Thiel E, Loffler H, et al: Prognostic factors in multicenter study for treatment of acute lymphoblastic leukemia in adults. Blood 71:123-131, 1988 32. Hussein KK, Dahlberg S, Head D, et al: Treatment of acute lymphoblastic leukemia in adults with intensive induction, consolidation, and maintenance chemotherapy. Blood 73:57-63, 1989 33. Radford JE Jr, Burns CP, Jones MP, et al: Adult acute lymphoblastic leukemia: Results of the Iowa HOP-L protocol. J Clin Oncol 7:58-66, 1989 34. Gaynor J, Chapman D, Little C, et al: A causespecific hazard rate analysis of prognostic factors among 199 adults with acute lymphoblastic leukemia: The Memorial Hospital experience since 1969. J Clin Oncol 6:1014-1030, 1988 35. Ellison RR, Holland JF, Weil M, et al: Arabinosyl cytosine: A useful agent in the treatment of acute leukemia in adults. Blood 32:507-523, 1968 36. Champlin R, Gale RP: Acute lymphoblastic leukemia: Recent advances in biology and therapy. Blood 73:20512066, 1989

37. Miller DR, Leiken S, Albo V, et al: Prognostic factors and therapy in acute lymphoblastic leukemia of childhood: CCG 141. Cancer 5:1041-1049, 1983 38. Crist W, Pullen J, Boyett J, et al: Acute lymphoid leukemia in adolescents: Clinical and biologic features predict a poor prognosis-A Pediatric Oncology Group study. J Clin Oncol 6:34-43, 1988 39. Bloomfield CD, Goldman AI, Alimena G, et al: Chromosomal abnormalities identify high-risk and low-risk patients with acute lymphoblastic leukemia. Blood 67:412420, 1986 40. Bloomfield CD, Wurster-Hill D, Peng G, et al: Prognostic significance of the Philadelphia chromosome in adult acute lymphoblastic leukemia, in Gale RP, Hoelzer D (eds): Acute Lymphoblastic Leukemia. UCLA Symposia on Molecular and Cellular Biology, New Series, Volume 108. New York, NY, Liss, 1990, pp 101-109 41. Sobol RE, Royston I, LeBien TW, et al: Adult acute lymphoblastic leukemia phenotypes defined by monoclonal antibodies. Blood 65:730-735, 1985 42. Kantarjian HM, Walters RS, Smith TL, et al: Identification of risk groups for development of central nervous system leukemia in adults with acute lymphocytic leukemia. Blood 72:1784-1789, 1988 43. Omura GA, Moffitt S, Vogler WR, et al: Combination chemotherapy of adult acute lymphoblastic leukemia with randomized nervous system prophylaxis. Blood 55:199203, 1980 44. Omura GA, Raney M: Long-term survival in adult acute lymphoblastic leukemia: Follow-up of a Southeastern Cancer Study Group trial. J Clin Oncol 3:1052-1058, 1985

Downloaded from ascopubs.org by University of Groningen on August 6, 2019 from 129.125.019.061 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.