[CANCER RESEARCH 58. 4173-4179. September 15. 1998]

Frequency of Prolonged Remission Duration after High-Dose Cytarabine

Intensification in Acute Myeloid Leukemia Variesby Cytogenetic Subtype1

Clara D. Bloomfield, David Lawrence, John C. Byrd,2 Andrew Carroll, Mark J. Pettenati, Ramana Tantravahi,

Shivanand R. Patii, Frederick R. Davey, Deborah T. Berg, Charles A. Schiffer, Diane C. Arthur, andRobert J. MayerThe Ohio State University. Columbus, Ohio 43210 ¡C.D. B.¡:Rom-ell Park Cancer Institute, Buffalo. New York 14263 ¡D.L.¡;Walter Reed Army Medical Center. Washington,D.C. 20307 ¡J.C. B.¡; University of Alabama at Birmingham, Birmingham. Alabama 35229 ¡A.C.I; Bowman Gray School of Medicine, Winston-Salem. North Carolina 27157¡M.J. P.I; Dana-Farber Cancer Institute. Boston. Massachusetts 02115 ¡R.T., D. T. B., R. J. M.¡;University of Iowa, Iowa City, Iowa 52242 [S. R. P.I; State University of NewYork Health Science Center at Syracuse, Syracuse, New York 13210 ¡F.R. D.¡;University of Maryland Cancer Center, Baltimore, Maryland 21201 ¡C.A. S.I; and University ofMinnesota. Minneapolis. Minnesota 55455 ¡D.C. A. I

ABSTRACT

Advances in the treatment of acute myeloid leukemia (AMD haveoccurred with the introduction of new therapies including high-dose

cytarabine and the identification of powerful prognostic factors such ascytogenetics that predict for long-term outcome. To date, the prognostic

impact of cytarabine dose escalation within various cytogenetic groups ofAML has not been assessed. We describe 285 newly diagnosed patientswith primary AML who had adequate karyotypes and were enrolled on aprospective Cancer and Leukemia Group B cytogenetic study. All patientswere randomly assigned to postremission treatment with standard-, intermediate-, or high-dose cytarabine intensification. Patients were catego

rized to one of three cytogenetic groups: (a) core binding factor type[(CBF); i.e., t(8;21) inv(16), t(16;16), and del(16)]; (¿>)normal; and (c)other abnormality karyotype. An evaluation of these patients after amedian follow-up time of over 7 years was performed to determine the

relationship of intensification to outcome by cytogenetic group. Patientsincluded 57 patients with CBF AML, 140 patients with normal karyotypeAML, and 88 patients with other cytogenetic abnormalities. The treatment outcome of CBF AML patients was superior, with an estimated 50%still in complete remission (CR) after 5 years as compared with 32 and15% for patients with normal karyotype AML and other abnormalityAML, respectively (P < 0.001). Univariate analysis showed the followingnonkaryotype factors to predict a prolonged CR duration: (a) younger age(P < 0.008); (b) lower leukocyte count (P = 0.01); (c) the presence of Auerrods (P = 0.004); «/) a lower percentage of bone marrow blasts(/' = 0.001) at the time of diagnosis, (e) and a higher postremission

cytarabine dose (P < 0.001). The impact of cytarabine dose on long-termremission was most marked (/' < 0.001) in the CBF AML group (after 5years, 78% of those with a dose of 3 g/m2 were still in CR, 57% of thosewith a dose of 400 mg/m2 were still in CR, and 16% of those with a doseof 100 mg/m2 were still in CR) followed by normal karyotype AML(P = 0.01; after 5 years, 40% of those with a dose of 3 g/m2 were still inCR, 37% of those with a dose of 400 mg/m2 were still in CR, and 20% ofthose with a dose of 100 mg/m2 were still in CR). In contrast, cytarabine

at all doses produced only a 21% or less chance of long-term continuous

CR for patients with other cytogenetic abnormalities. A multivariateanalysis of CR duration assessed the independent impact of each of thesevariables on cure. Significant factors entering this model in descending

Received 4/9/98: accepted 7/17/98.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore he hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1Supported by National Cancer Institute Grants 31946. 16058. and 26806 and by the

Coleman Leukemia Research Fund. C. D. B. was supported by Grants CA31946 andCA77658. D. L. was supported by Grant CA59518. J. C. B. was supported by GrantCA26806. A. C. was supported by Grant CA47545. M. J. P. was supported by GrantCA03927. R. T. and R. J. M. were supported by Grant CA3229I. S. R. P. was supportedby Grant CA47642. F. R. D. was supported by Grant CA2I060. C. A. S. was supported byGrant CA31983. D. C. A. was supported by Grant CA16450. This work was presented inpart at the Plenary Session of the December 1994 meeting of the American Society ofHemalology.

2 To whom requests for reprints should be addressed, at Hematology-Oncology Ser

vice. Ward 78. Walter Reed Army Medical Center. Washington. D.C. 20307. Phone: (202)782-5746; Fax: (202) 782-3256.

order of importance were cytogenetic group (CBF > normal > otherabnormality; P = 0.00001), cytarabine dose (3 g/m2 > 400 mg/m2 > 100mg/m2; /' = 0.00001), logarithm of leukocyte count at the time of diagnosis

(P = 0.0005), and histological subtype of AML (P = 0.005). This study

demonstrates that the curative impact of cytarabine intensification variessignificantly among cytogenetic groups and results in a substantial prolongation of CR among patients with CBF and normal karyotypes, but notin those with other karyotypic abnormalities. These findings support theuse of pretreatment cytogenetics in risk stratification of postremissionAML therapy.

INTRODUCTION

Advances in understanding the pathogenesis of AML and improvement in the treatment of AML1 have occurred during the last two

decades. In particular, cytogenetics has become one of the single mostuseful factors for predicting outcome to both induction and postremission therapy (1-4). Specifically, adult patients presenting with pri

mary AML who possess favorable cytogenetic abnormalities [i.e.,t(8;21) or inv(16) karyotypes] have greater than an 80% chance ofattaining a CR and generally have an improved disease-free survival

when compared with that of other cytogenetic groups (5, 6). It hasrecently been determined that the gene transcriptional products ofboth t(8;21) and inv(16) lead to structural alterations of differentcomponents of CBF. which is involved in the regulation of hemato-poiesis (7-9). Thus, although different chromosomes (chromosomes 8

and 21 or 16) are involved, the t(8;21) and inv(16) karyotypes arecommonly referred to as CBF AML. In contrast, patients with othercytogenetic abnormalities involving deletions of chromosome 5 or 7,other unbalanced translocations, or primary chromosome additions(trisomy 4, 8, 11, or 13) have inferior remission rates, remissiondurations, and overall survival. (1-5) The majority of reported data

documenting both favorable and unfavorable treatment outcome inparticular cytogenetic subsets has been derived from small case seriescharacterized by variable and often unspecified postremission treatment regimens. Additionally, no randomized treatment study in adultAML to date has assessed the impact of dose intensification of anyactive agent on specific cytogenetic groups.

The introduction of postremission strategies such as high-dose

cytarabine and allogeneic and autologous stem cell transplantation hascontributed to an improved outcome for adult patients with AML. Forhigh-dose cytarabine, the steep in vitro dose-response curve observed

with this agent (10), combined with initial small studies (11, 12),prompted large randomized studies (13-16) that established high-dose

cytarabine intensification therapy as a commonly accepted treatmentfor young patients with AML. Allogeneic and autologous stem cell

3 The abbreviations used are: AML. acule myeloid leukemia: CALGB. Cancer and

Leukemia Group B: CBF. core binding factor type; CR. complete remission: FAB.French-American-British.

4173

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

transplantation have similarly been demonstrated to have efficacy inthe initial management of AML but have higher treatment-relatedmortality rates (17-19). Randomized studies comparing these therapies with high-dose cytarabine have failed to demonstrate any survivalbenefit (17-19). Thus, debate persists as to the optimal therapy for

young patients with AML. Utilization of pretreatment cytogenetics tostratify patients to particular treatments has been advocated, but therelationship of cytogenetic subtype and the dose intensity of cytarabine as it relates to the long-term remission of patients with AML has

not been reported. Herein, we describe such an analysis of outcome ofcytogenetically defined groups of patients who were assigned toreceive varying doses of postremission cytarabine therapy.

PATIENTS AND METHODS

Patients. Patients included in this analysis were 16 years or older withprimary AML as defined by the FAB classification system (20) and wereparticipants in a previously reported randomized trial conducted by theCALGB (13). Patients with a prior history of myelodysplasia. other antecedenthematological malignancies, prior nonsteroidal cytotoxic chemotherapy orradiation therapy, preexisting liver disease as defined previously (13), oruncontrolled infection were excluded. Central review of the pathologicaldiagnosis was performed. Due to the recently recognized unique biology ofpatients with acute promyelocytic leukemia, such patients have been excludedfrom this analysis of postremission therapy.

Cytogenetics. Chromosomal analysis of bone marrow was performed ininstitutional CALGB cytogenetics laboratories, and karyotypes were centrallyreviewed biannually by an expert panel of CALGB cancer cytogeneticists.Specimens were obtained at diagnosis from all patients. Specimens wereprocessed using direct methods and unstimulated short-term (24, 48, and 72 h)cultures. G-banding was usually done, although Q-banding was acceptable for

inclusion in this series. A minimum of 20 bone marrow metaphase cells wasanalyzed in each patient designated as having a normal karyotype. The criteriaused to describe a cytogenetic clone and the description of karyotype followedthe recommendations of the International System for Human CytogeneticNomenclature (21).

Treatment. The therapy administered to patients included in this study hasbeen described previously (13). Briefly, the patients received induction therapyconsisting of 7 days of a continuous infusion of cytarabine (200 mg/m2/day)

and 3 days of bolus daunorubicin (45 mg/nr/day for patients s 60 years and30 mg/nr/day for patients > 60 years). Patients not attaining a CR received a

second course of induction therapy consisting of 5 days of cytarabine and 2days of daunorubicin at doses identical to those administered initially. Patientsin CR were stratified according to the number of induction courses (one ortwo) and their age (less than 40 years, 40-60 years, and more than 60 years)

and randomly assigned by the CALGB statistical center to receive four coursesof single-agent cytarabine in one of three regimens: (a) 100 mg/m2/day for 5days as a continuous infusion; (b) 400 mg/m2/day for 5 days as a continuousinfusion; and (c) 3 g/m2 bolus over 3 h every 12 h on days 1.3, and 5 for a total

of six doses. These courses were administered approximately every 28 days,unless recovery from toxicity had not occurred. After cytarabine, patientsreceived four monthly treatments with 100 mg/m2 cytarabine every 12 h by s.c.injection for 5 days (total of 10 doses) and 45 mg/m2 daunorubicin by rapid

infusion on day 1.Evaluation. During treatment, patients underwent a bone marrow aspira

tion after the completion of each cytarabine intensification and maintenancetherapy. Thereafter, patients were followed with bone marrow testing every 3months for 1 year, every 6 months for 2 years, and then every year for 2additional years. Patients were followed yearly after 5 years of remission withbone marrow examinations being performed only if the blood counts suggesteda relapse of AML.

Criteria for Response and Definition of Relapse. A CR was defined asthe presence of a morphologically normal bone marrow and at least 1,500granulocytes//il and 100,000 platelets//^.! in the blood. Relapse was defined asgreater than 5% leukemic cells in bone marrow aspirates or new extramedul-

lary leukemia in patients with a previously documented CR using the previously published National Cancer Institute guidelines (22).

Statistical Analysis. The level of significance was evaluated with two-

sided tests. Uncensored continuous variables and percentages were comparedamong groups using the Kruskal-Wallis test. Nominal (discrete) variables werecompared among groups with Pearson's x1 test. The Kaplan-Meier method

was used to estimate the distribution of survival and CR duration. Censoreddata were compared among groups using the log-rank test, Cox's proportional

hazards model, or the Tarone-Ware trend test (23). Pairwise comparisons were

performed as appropriate.An intent-to-treat analysis was performed based on randomization to one of

the three intensification arms after CR. Survival end points, measured from thedate of randomization to that of intensification, were death (failure) and aliveat last follow-up (censored). For CR duration, as measured from the date of

randomization, the end point was the first occurrence of relapse (failure), deathin CR (censored), bone marrow transplant while in CR (censored), or continuous CR at last follow-up (censored). The three cytogenetic groups of interest

were: (a) group CBF [CBF karyotypes i.e., t(8;21)(q22;q22), inv(16)(pl3q22),t(16;16)(pl3;q22), and del(16)]; (b) group NL (normal karyotype); and (c)group other (other karyotypes).

The median follow-up time was 92 months for the 96 patients censored for

survival and 86 months for the 88 censored patients for CR duration. Of thelatter 88 censored patients. 12 died during the first CR, and 4 were transplantedwhile in the first CR. Because substantial censoring occurred for CR durationbeyond 48 months, an analysis of the Kaplan-Meier plateau for cure was

performed with the test of Mailer and Zhou (24). In this regard, 37 cases hadCR durations longer than the last relapse at 91.4 months. Follow-up was

lengthy enough on this plateau (P < 0.001) to indicate that a cure in thestatistical sense (P < 0.001) had occurred, and a cure rate model would beapplicable (25). At a median follow-up of over 7 years for 88 censored cases,

there is strong evidence that modeling for cure is justified. There were onlyfive relapses beyond 5 years and two relapses beyond 7 years of follow-up,

respectively.For multivariable analysis of CR duration, the Farewell cure rate model, a

parametric mixture model, was selected (26, 27). A Weibull function approximated the relapse process, and the other term described the cure. The Weibullparametric regression model includes covariates only in the scale (not in shape

or term) and therefore has the proportional hazard property (28). A logisticmodel with covariates determined the proportion labeled as cured. An initialscreening of variables was made with either the log-rank test or the propor

tional hazards model at a P of 0.1 (29). Using the 264 cases complete for allcandidate variables, location in the model at either the logistic or Weilbullcomponent was checked separately for each variable (P = 0.1 level, likelihood

ratio test) for continued testing during the model building. Indicator variableswere used for discrete variables. All covariates were centered. Variables wereentered with forward selection into the model using a likelihood ratio test untilno variable entered at the P = 0.01 level. The interaction of cytogenetics with

intensification arm was examined and did not enter the model. Fit wasevaluated graphically.

RESULTS

Validation of Using Cytogenetically Defined Patients on theCALGB Treatment Study. Because this report includes only thesubset of patients enrolled on the CALGB therapeutic trial who hadadequate, pretreatment, centrally reviewed cytogenetics, a comparisonof outcome between this subset and the remaining patients wasperformed. A total of 1104 patients with AML were enrolled on theprotocol between October 31, 1985 and October 1, 1990, of whom1088 were eligible. Of these 1088 patients, 826 were enrolled on theprospective cytogenetic study. Inadequate or incomplete cytogenetics(n = 216) were documented on the central karyotype review, leaving

610 patients with adequate pretreatment cytogenetics. Of the 610patients having adequate cytogenetics, 19 myelodysplasia cases(based on central morphological review) and 66 acute promyelocyticleukemia cases were excluded. Of the remaining 525 patients, 321(61%) attained a CR, with 285 (89%) subsequently being randomizedto postremission cytarabine. Thus, 285 patients were évaluableforassessment of the dual impact of cytogenetics and cytarabine dose

4174

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

intensity on prolonging the CR duration of AML. To assure thatpatients with adequate cytogenetic studies were representative of theentire cohort of AML patients, an analysis of outcome among thesethree groups of patients was performed [i.e., patients who wereenrolled on the cytogenetic study with adequate cytogenetics (knowncytogenetic subtype), patients who were enrolled on the cytogeneticstudy without adequate cytogenetics (unknown cytogenetic subtype),and patients who were not enrolled on the cytogenetic study (unknown cytogenetic subtype)]. The CR rate, CR duration, and overallsurvival as defined by CALGB conventions (30) demonstrated nosignificant differences among the groups (data not shown), therebyjustifying the cytogenetic group of patients with adequate karyotypedata as being representative of all patients enrolled on the treatmentstudy.

Preintensification Patient Features. Of the 285 randomizedpatients, 57 patients were classified as group CBF [CBF karyotypes,i.e., t(8;21)(q22;q22), inv(16)(pl3q22), t(16;16)(pl3;q22), anddel(16)(q22)], 140 patients were classified as group NL (normalkaryotype), and 88 patients were classified as group other (otherkaryotypes). Group CBF included patients with two biologically similar (i.e., t(8;21) and rearrangements of 16q22) chromosomal abnormalities that had statistically similar outcomes in terms of CR rate, CRduration, and overall survival. Group other included a wide variety of

Table 1 Characteristics of randomized CR patients with adequate cytogenetics

Clinical features"Group CBF Group NL Group other(n = 57)* (n = \40)h (n = 88)fc

Age(yr)''%

patients >60yr%female%hepatomegaly%splenomegaly%

extramedullaryleukemiaLeukocytecount X109/liter'Platelets

X109/literrHemoglobing/dl'Auer

rods present(%)'%peripheral bloodblasts'%bone marrowblasts'FAB

classMl(%)M2(%)M4(%)M5

(%)Other(%)%

CR with oneinductioncycle351253571922.2448.575526025439248950294968619.3599.558567120392116474482649910721.5589.5466078262615201372<0.0010.050.870.620.740.010.540.030.020.0020.54<0.001<0.0010.03

" Ali pretreatment, except for the induction cycles required for CR.* See the text for definitions of the cytogenetic groups.' Median.

P < 0.001

CBF AML (n=57)i in i| ^—inin i i i

Normal Karyotype AML (n=140)

ninni i—m-

Other Karyptype AML (n=88)

4 6 8 10 12Years

1.0O)

| 0.81¿g 0.6

| 0.4oo 0.2

Q_

0.0

Normal Karyotype AML (n=140)

P < 0.001

AML (n=57)niuun ni i ii i

Other Karyotype AML (n=88)

02468Years

Fig. 2. Survival by cytogenetic group.

10 12

Table 2 Univariate analysis of selected factors affecting CR duration

MedianNo.of continuous CR %5-yrClinical

featureCytogenetics"-*Group

CBFGroupNLGroup

OtherCytarabinedose3g/m2400

mg/m-100mg/m2Bone

marrow blasts(%)"Auerrods"PresentAbsentAge

(yr)°«60a61Leukocyte

count(logarithm)"FAB"MlM2M4M5Other^No.

of inductioncycles12%

peripheral bloodblasts""

Pretreatmentvalues.bCoa tho tovt fnr 1 .Õ 1 t •patients5714088909996278163119215702835210764401921668260e.. „...,.„,(mo)36.917.510.521.817.511.9C17J12.516.815.3C13.018.017.111.110.115.815.8Ccontinuous

CR503215423317C36223517C17373124263420CP<0.001<0.0010.00

10.0040.0080.010.030.0530.08

Fig. 1. CR duration by cytogenetic group.

c Analyzed as a continuous variable.d M6, M7. MO, undifferentiated, mixed, or unclassifiable.

abnormalities including isolated trisomy 8 (n = 10), Ilq23 abnormalities (n = 13), other isolated trisomies (n = 9), 5q deletions(n = 4), other single structural abnormalities (n = 38), other dualabnormalities (n = 8), and complex karyotypes (n = 6). The fre

quency of CR patients randomized to intensification therapy in groupCBF (93%), group NL (89%), and group other (86%) was similar(P = 0.37). Similarly, there was no difference in the distribution of the

three cytarabine dose schedules among the three cytogenetic groups(P = 1.0). The pretreatment clinical features of patients in cytogenetic

groups CBF, NL, and other randomized to receive intensification aresummarized in Table 1. Patients with group CBF karyotypes wereyounger and had a lower percentage of bone marrow blasts as compared with group NL and group other karyotypes (P < 0.001). Ofnote, only three patients with normal cytogenetics and one withanother abnormality were classified with a M4Eo FAB histologicalsubtype of AML.

Clinical Outcome by Cytogenetic Group. An overall differencein CR duration (Fig. 1) and survival (Fig. 2) was noted among

4175

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

1.0

g °-8•l 0.6

.0'"§ 0.4

Q.

£L 0.2

0.0

P < 0.001

3gm/m (n=90)

'in i n—ii u i i i400mg/m2 (n=99)

-w-100mg/m (m=96)

0 468Years

10 12

Fig. 3. CR duration by cytarabine dose intensification

karyotype groups CBF, NL, and other (P < 0.001). The estimatedpercentage of patients remaining in CR at 5 years was 50% for groupCBF, 32% for group NL, and 15% for group other, respectively(Table 2). Compared with group other patients, paired comparisons ofCR duration demonstrated a difference for group CBF (P < 0.001)and group NL (P = 0.001) patients. Because this analysis used anintent-to-treat design, completion of all four intensification therapies

by each cytogenetic group could partially explain these results. Therefore, we examined the frequency of completion of intensificationwithin the three cytogenetic groups. The frequency of completingeach dose of cytarabine therapy was statistically similar among eachof the cytogenetic groups. Furthermore, the great majority (93%) ofpatients completing intensification therapy in this treatment studywent on to receive the assigned maintenance therapy, making this anunlikely source for explaining the difference in CR duration amongthe different cytogenetic groups. Similarly, the difference in outcomeof different cytogenetic groups was not significantly altered by thesmall percentage (6%) of patients who were censored for death in thefirst CR, with a similar representation in group CBF (7%), group NL(6%), and group other (3%).

Patients in group CBF also had a prolonged survival. The estimatedpercentage of patients surviving at 5 years was 64% for group CBF ascompared with 35 and 27% in group NL and group other, respectively. Paired comparisons demonstrated a difference in survival between group CBF patients as compared with both group NL(P = 0.002) and group other (P < 0.001) patients. However, survival

was not different between group NL and group other patients(P = 0.10).

Univariate Analysis of Factors Affecting CR Duration. In anattempt to define factors other than the cytogenetic subtype of AMLthat might impact on CR duration, a univariate analysis of a numberof pretreatment features along with assignment of cytarabine intensification dose was performed and is summarized in Table 2. In additionto cytogenetic subtype, age ^ 60 years, a lower leukocyte count at thetime of diagnosis, the presence of Auer rods in the myeloblasts, alower percentage of bone marrow blasts, and a higher postremissioncytarabine dose were each associated with a longer CR duration. Ashas been reported previously (13), CR duration for all patients wasrelated to the cytarabine dose (Fig. 3). The estimated proportion ofpatients remaining in CR at 5 years by treatment group was 42% forthe high-dose cytarabine group, 33% for the intermediate-dose cytarabine treatment group, and 17% for the standard-dose cytarabine

treatment group, respectively (P < 0.001). Only 16 patients (6%)were censored for death during the first CR, with the frequency beingsimilar among the high-dose (6%), intermediate-dose (6%), and low-

dose (5%) cytarabine arm. Paired comparisons demonstrated a markeddifference in CR duration between patients receiving either high-dose(P < 0.001 ) or intermediate-dose (P = 0.003) cytarabine as comparedwith those receiving standard-dose cytarabine. However, CR durationwas not different among patients receiving high-dose cytarabineas compared with those receiving intermediate-dose cytarabine(P = 0.63).

Assessment of the response duration to various intensificationtherapies by cytogenetic subtype of AML is summarized in Table 3.A trend test of intensification level stratified by cytogenetic groupsrevealed an overall difference (P < 0.001). In particular, with subsetanalysis, it is notable that the remission duration of patients with theCBF karyotype increases (P < 0.001) with an increasing dose ofcytarabine intensification (Fig 4a). Patients with CBF karyotypesreceiving 100 mg/m2 cytarabine intensification had only a 16%

chance of remaining in CR at 5 years as compared with a 78%frequency in those receiving 3 g/m2 cytarabine. The benefit of cyt

arabine intensification was less obvious in patients with normal karyotypes (P = 0.01) and was absent in patients with other cytogeneticabnormalities (P = 0.10; Fig. 4, b and c, respectively).

Multivariate Analysis to Assess Characteristics Impacting onthe Cure of AML. Because the median follow-up after attaining CR

in this cohort of patients is in excess of 7 years with a prolongedplateau, the use of a multivariate analysis that addresses the impact ofdifferent variables on both potential cure and treatment failure wasdeemed to be statistically appropriate. Candidate pretreatment clinicalfeatures including cytogenetic types of AML, age (60 years or less and61 years or greater), the percentage of both blood and bone marrowblasts at the time of diagnosis, the log of the leukocyte count, histo-

Table 3 CR duration by cytogenetic group according to cytarabine dose randomization

Cytogeneticgroup"Group

CBFGroup

NLGroup

otherCytarabine

dose3

g/m"400mg/m2100

mg/m23

g/m2400mg/m2100

mg/m23

g/m2400mg/m2100

mg/m~No.

ofpatients182019454847273130Median

time of CR(mo)NRdNRJ14.318.221.412.513.310.69.6%

5-yr CR estimate(95%Ci)*78

(59-97)57(34-80)16(0-32)40

(25-54)37(13-51)20

(8-32)21

(5-37)13(1-25)13(0-26)%

cureestimater66522347321217103

0 See the text for definitions of the different cytogenetic groups.

Kaplan-Meier estimates. Cl. confidence interval.' Estimates were obtained using Farewell mixture model.J NR. not reached.

4176

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

trüC

Co'•coCLO

B

ocoeCo'•eoQ.

gCL

CCÜ

g'•CoQ_o

1.0

0.8

I 0.6

0.4

O- 0.2

0.0O

1.0

0.8

- 0.6

0.4

0.2

0.0O

P = 0.0004

3gm/m¿-i—H—i-

(n=18)

-H—I- I II l l

400mg/m (n=20)

100mg/m' (n=19)

6Years

8 10 12

P = 0.012

3gm/m (n=45)*—i—HUÕ I-H

'III I II H

400mg/m (n=48)

100mg/m (n=47)

6Years

8 10 12

P = 0.10

3gm/m (n=27)—H 1—l—h

100mg/m2(n=30),400mg/m2- (n=31)

6Years

8 10 12

Fig. 4. CR duration of patients within specific groups by cytarabine dose intensification. A, group CBF; B, group NL; C. group other.

logical type (i.e., FAB classification Ml, M2, M4, M5, and other), andthe presence or absence of Auer rods as well as the number ofinduction cycles needed to achieve CR and the postinduction cytarabine dose were tested in the model based on the univariate analysisdescribed previously. Of these factors, cytogenetic subtype (group

CBF > group NL > group other; P = 0.00001) entered the cure

model first, followed by the dose of cytarabine intensification (3g/m2 > 400 mg/m2 > 100 mg/m2; P = 0.00001) as summarized in

Table 4. Two other variables entered the model: (a) the log of theleukocyte count (P = 0.0005); and (b) the FAB subtype (P = 0.005),

with a lower leukocyte count and FAB classification M4(M4 > M2 > M5 > Ml > other FAB classifications) predicting forthe later relapse of AML. Thus, cytogenetic group and intensificationlevel estimate the proportion potentially cured, whereas the FABsubtype and the log of the leukocyte count are prognostic factors forrelapse using this statistical cure model.

DISCUSSION

Our data demonstrate that both pretreatment karyotype and high-

dose cytarabine intensification independently impact the chance ofattaining long-term remission and apparent cure of adult AML. Randomized treatment studies (13-15) examining high-dose cytarabine

dose intensification, a therapy currently considered standard foryounger patients with AML, have not included an assessment of thecompeting impact of both cytarabine dose and cytogenetics on thepotential cure of AML. Other factors such as age, the absence of Auerrods, and extramedullary leukemia, which have been identified asprognostic factors in earlier AML studies, were also examined in thisstudy and were found not to independently contribute to the likelihoodof AML cure. Indeed, it is well documented that adverse risk factorssuch as older age and the absence of Auer rods occur more frequentlyin poor-risk cytogenetic groups (1-5, 31-33). These adverse riskfeatures are underrepresented in most series with CBF AML (34-37).The coassociation of poor-risk clinical features with an adverse cytogenetic aberration or of a better-risk clinical feature with CBF AML

renders the interpretation of outcome impossible unless the relationship of all variables is considered. Our multivariate analysis assessingindependent variables that predict for statistical cure demonstratedthat only the cytogenetic type of AML and the cytarabine postremis-sion dose seem to be major prognostic factors for long-term remission

and apparent cure of adult patients with AML. Importantly, unlikemost AML treatment studies that have relied on models to projectlong-term estimates of disease-free outcome beyond 5 years, our studyhas a >7-year median patient follow-up of censored cases. Thislong-term follow-up lends credence to the model for cure used in this

study.An equally important finding of this study is that increasing the

intensity of cytarabine is not of significant benefit to all cytogeneticgroups. Dose escalation of cytarabine in patients with CBF karyotypesand, to a lesser extent, in patients normal karyotypes significantlyincreases the duration of prolonged remission and likely cure. Such a

Table 4 Multicanale analysis of factors impacting potential cure of AML

FactorCytogenetics"Group

CBFGroupNLGroup

otherCytarabineintensification3g/m2400mg/m2100

mg/m2Regression

coefficient0.823-1.4571.8741.269CSEofthemean0.3660.4560.4340.437P0.000010.00001

" See the text for definitions of the cytogenetic groups.h Positive regression coefficient indicates an improved cure rate for CBF as compared

with the normal (NL) cytogenetic AML group; negative regression coefficient indicates alower cure in group other as compared with the NL AML cytogenetic group.

c Increasing regression coefficient indicates a higher cure rate with dose escalation of

cytarabine intensification relative to the 100 mg/m group. FAB classification and the logof the leukocyte count also entered the model in the relapse portion.

4177

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

10.

cytogenetically specific dose escalation finding has not been identified previously in any randomized adult AML trial. However, othernonrandomized studies have reported favorable results of high-dose

cytarabine in patients with CBF karyotypes as compared with othercytogenetic groups, thus supporting these findings (38). The reasonsfor this differential response are uncertain but likely include molecularaberrations that render cells with CBF karyotypes more sensitive tohigh-dose cytarabine. Additional laboratory studies investigating thebiological basis for CBF AML cell sensitivity to high-dose cytarabine

are needed.One limitation of this study is the narrow eligibility criteria that

only allowed AML patients without antecedent hematological disorders or cytopenias to be enrolled on the treatment study. Theseexclusion criteria limit the generalizability of our results to a largeproportion of patients with AML seen in the community who wouldhave been ineligible for this study. Patients with refractory anemiawith excess blasts or refractory anemia with excess blasts in transformation were also ineligible for the treatment study. Consequentially,individuals later identified as having these findings on central reviewwere excluded in the current study. This later group was excluded inthe current study to maintain a homogeneous group of patients,although the frequency of myelodysplastic syndrome patients excluded on central review among the different cytogenetic groups wassimilar. The exclusion of this small number of patients did not alterthe conclusions of this study.

The ramifications of our study, as it relates to the current management of adult AML patients, are substantial. In the setting of cytogenetic risk stratification for postremission intensification, our datastrongly support the use of high-dose cytarabine as the best treatment

option for postremission therapy in patients with CBF AML. For thehighly heterogeneous group of AML patients with a normal karyo-

type, the choice of postremission therapy is less certain and willrequire the identification of cytogenetically undetectable molecularaberrations, such as the ALL-1 duplication, that predict for pooroutcome (39). Randomized studies comparing the efficacy of high-

dose cytarabine with new novel intensification therapies in suchpoor-risk patients with normal karyotypes seem warranted. Although

there may be certain other karyotypic abnormalities that benefit fromhigh-dose cytarabine, the current data suggest that these patients

should be treated with novel therapeutic approaches with either newtherapeutic agents alone or possibly in concert with allogeneic transplant. Based on the data presented, CALGB currently is embarking ona cytogenetically stratified treatment approach to adult patients withAML.

REFERENCES

1. Arthur, D. C, Berger. R., Colomb. H. M., Swansbury, G. J.. Reeves, B. R„Alimena,G., Van Den Berghe. H., Bloomfield, C. D., de la Chapelle, A.. Dewald, G. W.,Garson, O. M., Hagemeijer, A., Kaneko, Y.. Mitelman, F., Pierre, R. V., Ruutu, T..Sakurai. M.. Lawler, S. D.. and Rowley. J. D. The clinical significance of karyotype ^2.in acute myelogenous leukemia. Cancer Genet. Cytogenet., 40: 203-216, 1989.

2. Dastugue. N., Payen, C., Lafage-Pochitaloff, M., Bernard. P.. Leroux, D., Huguet-Rigal, F., Stoppa, A-M., Marii. G.. Molina, L., Michaile!, M., Maraninchi. D., Altai.M., and Reiffers, J. Prognostic significance of karyotype in de novo adult acutemyelogenous leukemia. Leukemia (Baltimore), 9: 1491-1498. 1995. 23.

3. Stasi, R., Del Poeta. G., Masi. M., Tribalto. M., Venditti. A.. Papa. G.. Nicoletti, B.,Vernole, P., Tedeschi, B., Delaroche, I., Mingarelli, R., and Dallapiccola, B. Inci- 24.dence of chromosome abnormalities and clinical significance of karyotype in de novoacute myeloid leukemia. Cancer Genet. Cytogenet., 167: 28-34, 1993. 25.

4. Schiffer. C. A.. Lee, E. J., Tomiyasu, T., Wiernik, P.. and Testa. J. R. Prognosticimpact of cytogenetic abnormalities in patients with tie novo acute nonlymphocytic 26.leukemia. Blood. 73: 263-270. 1989.

5. Mrozek. K., Heinonen. K.. de la Chapelle. A., and Bloomfield. C. D. Clinical 27.significance of cytogenetics in acute myeloid leukemia. Semin. Oncol., 24: 17-31.1997. 28.

6. Larson, R. A., Williams, S. F., Le Beau, M. M., Bitter, M. A., Vardiman, J. W., andRowley. J. D. Acute myelomonocytic leukemia with abnormal eosinophils and 29.inv(I6) or t(I6:l6) has a favorable prognosis. Blood. 68: 1242-1249, 1986.

4178

17.

18.

19.

20.

21-

Miyoshi, H.. Shimizu. K.. Kozu, T., Maseki. N.. Kaneko. Y., and Ohki. M. Thet(8;21) breakpoints on chromosome 21 in acute myeloid leukemia clustered within alimited region of a novel gene AML/. Proc. Nati. Acad. Sci. USA. 88: 10431-10434,

1991.Liu, P., Tarle. S. A., Hajra. A.. Claxton, D. F., Mariton. P.. Freedman. M.. Siciliano,M. J., and Collins, F. S. Fusion between transcription factor CBFß/PEBP2/3and amyosin heavy chain in acute myeloid leukemia. Science (Washington DC), 267:1041-1044, 1993.

van der Reijden, B. A.. Bloomfield, C. D., Touw, I. P.. and Jansen, J. H. Acuteleukemias with structurally altered core binding factor subunits (t(8::21), inv(16).t(12;22)). Leukemia (Baltimore), /A- 2217-2219, 1997.

Skipper, H. E., Schaebel, F. M.. Jr.. and Wilcox, W. S. Experimental evaluation ofpotential anticancer agents. XIII. On the criteria and kinetics associated with "curability" of experimental leukemia. Cancer Chemother. Rep.. 35: 1-11, 1964.

Champlin, R.. Gajewski. J., Nimer. S., Vollset. S., Landaw. E.. Winston, D.. Schiller.G., and Ho. W. Postremission chemotherapy for adults with acute myelogenousleukemia: improved survival with high-dose cytarabine and daunorubicin consolidation treatment. J. Clin. Oncol., 7: 1199-1206, 1990.Wolff, S. N., Herzig. R. H., Fay, J. W.. Phillips. G. L., Lazarus, H. M.. Flexner, J. M.,Stein. R. S.. Gréer,J. P., Cooper. B.. and Herzig. G. P. High-dose cytarabine and

daunorubicin as consolidation therapy for acute myeloid leukemia in first remission:long-term follow-up and results. J. Clin. Oncol., 7: 1260-1267, 1989.

Mayer. R. J.. Davis, R. B., Schiffer, C. A., Berg, D. T.. Powell. B. L.. Schulman, P..Omura, G. A., Moore, J. O.. Mclntyre, O. R.. and Frei. E. Intensive postremissionchemotherapy in adults with acute myeloid leukemia. N. Engl. J. Med., 331: 896-

903, 1994.Bishop, J. F., Matthews, J. P.. Young, G. A., Szer, J., Gillett, A., Joshua. D.,Bradstock. K., Enno, A., Wolf, M. M., Fox, R.. Cobcroft, R.. Herrmann, R.. Van DerWeyden, M.. Lowenthal. R. M., Page, F., Garson, O. M., and Juneja. S. A randomizedstudy of high-dose cytarabine in induction of acute leukemia. Blood, 87: 1710-1717,

1996.Weick, J. K., Kopecky, K. J., Appelbaum, F. R., Head, D. R.. Kingsbury. L. L.,Balcerzak, S. P., Bickers, J. N., Hynes, H. E., Welborn, J. L., Simon. S. R.. andGrever. M. R. A randomized investigation of high-dose versus standard-dose 1-ß-D-

arabinofuranosylcytosine with daunorubicin in patients with previously untreatedacute myeloid leukemia: a Southwest Oncology Group Study. Blood. 88: 2841-2851.

1996.Büchner,T.. Hiddemann, W., Löffler,G.. Gassmann, W.. Maschmeyer, G., Heit, W..Hossfeld. D.. Weh, H., Ludwig, W. D.. Thiel, E., Nowrousian, M.. Aul, C., Lengfelder, E., I alluni. B.. Mainzer. K.. Urbanitz. D., Emmerich. B.. Middelhoff. G.,Donhuijsen-Ant, H. R., Hellriegel, H. P., and Heinecke, A. Improved cure rate by

very early intensification combined with prolonged maintenance chemotherapy inpatients with acute myeloid leukemia: data from the AML cooperative group. Semin.Hematol., 28: 76-79, 1991.

Zittoun, R. A., Mandelli. F., Willemze, R., Witte, T. D., Labar. B.. Resegotti. L.,Leoni, F.. Damasio, E., Visani, G., Papa, G., Caronia, F., Hayal, M., Stryckmann. P.,Rololi, B., Leoni, P.. Peelermans, M. E., Dardenne, M., Vegna, M. L., Petti, M. C.,Solbu, G.. and Suciu. S. Autologous or allogeneic bone marrow transplant comparedwith intensive chemotherapy in acute myeloid leukemia. N. Engl. J. Med., 332:217-223, 1995.

Harousseau, J. L., Cahn, J. Y.. Pignon. B., Wilz, F., Delain, M.. Lioure, B.. Lamy. T.,Desablens. B., Briere, J., Cassasus, P., Audnuy, B., Polin, V., Tellier, Z., andHurteloup, P. Aulologous bone marrow transplantalion versus intensive consolidationchemotherapy as post-remission therapy in aduli acule myeloblaslic leukemia: finalanalysis of Ihe Goelamm randomized sludy. Proc. Am. Soc. Clin. Oncol., /5.- 358,

1996.Cassilelh, P., Harringlon, D.. Paiella, E., Lazarus, H., Appelbaum, F., Bennell, K. J.,Hurd. D., Willman, C., Magenis, E., and Wiernik, P. Comparison of autologous bonemarrow transplant with high-dose cylarabine in acule myeloid leukemia in firsl

remission: an ECOG inlergroup study. Proc. Am. Soc. Clin. Oncol., 16: 89, 1997.Bennell, J. M., Calovsky, D.. Daniel, M. T., Flandrin. G.. Gallon, D. A. G., Gralnick,H. R., and Sultan, C. Proposed revised criteria for Ihe classificaiion of acute myeloidleukemia: a report of the French-American-British Cooperative Group. Ann. Intern.Med., 103: 626-629. 1985.

ISCN. An International System for Human Cytogenelic Nomenclalure. Basel: S.Karger, 1995.Cheson, B. D., Cassilelh, P. A., Head, D. R., Schiffer, C. A., Bennett. J. M..Bloomfield, C. D., Brunning, R., Gale. R. P., Grever. M. R., Keating, M. J., Sawilsky,A.. Slass, S., Weinslein, H., and Woods. W. Report of Ihe Nalional Cancer Inslitute-sponsored workshop on definitions of diagnosis and response in acule myelogenousleukemia. J. Clin. Oncol., 8: 813-819, 1990.Tarane, R. E., and Ware, J. On dislribution-free tests for equalily of survivaldistributions. Biometrika. 64: 156-160, 1977.Mailer. R. A., and Zhou, S. Eslimaling the proportion of immune in a censoredsample. Biomelrika, 79: 731-739, 1992.Goldman, A. Survivorship analysis when cure is a possibilily: a Monle Carlo study.Stai. Med., 3: 153-163, 1984.

Farewell. V. T. The use of mixture models for the analysis of survival data wilhlong-lerm survivors. Biomelrics. 38: 1041-1046, 1982.Farewell, V. T. A model for a binary variable with time-censored observations.Biometrika. 64: 43-46, 1977.Lawless, J. F. Slalislical Models and Melhod for Lifelime Dala. New York: JohnWiley and Sons, 1982.Halpern, J.. and Brown, B.W. Cure rale models: power of the logrank and generalizedWilcoxon lesls. Stai. Med., 6: 483-489, 1987.

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

REMISSION DURATION AFTER HIGH-DOSE CYTARABINE IN AML

30. Moore, J. O.. Dodge, R. K.. Amrein. P. C.. Kolilz, J.. George, S.. Dodge, R., Rizzieri,D. A., and Schiffer. C. A. Granulocyte-colony stimulating factor accelerates granu-

locyte recovery after intensive postremission chemotherapy for acute myeloid leukemia with a/.iridinyl benzoquinone and mitoxantrone: Cancer and Leukemia Group BStudy 9022. Blood, 89: 780-788, 1997.

31. Cuneo. A.. Ferrant, A.. Michaux. J. L., Boogaerts. M., Demuynck, H., Van Orshoven.A., Criel. A.. Stul, M., Cin, P. D.. Hernandez. J.. Chatelein, B., Doyen. C.. Louwagie.A., Castoldi, G.. Cassiman. J. J., and Van Den Berghe, H. Cytogenetic profile ofminimally differentiated (FAB MO) acute myeloid leukemia: correlation with clini-cobiologic findings. Blood. 85: 3688-3694. 1995.

32. Stasi. R.. Del Poeta, G.. Venditti. A.. Masi, M.. Stipa, R., Dentamaro. T., Cox, C.,Dallapiccola, B., and Papa. G. Analysis of treatment failure in patients withminimally differentiated acute myeloid leukemia (AML-MO). Blood. 83: 1619-

1625. 1994.33. Rowley, J. D., Alimena. G.. Garson, O. M.. Hagemeijer, A.. Mitelman, F., and

Prigogina, E. L. Collaborative study of the relationship of the morphological type ofacute nonlymphocytic leukemia with patient age and karyotype. Blood, 59: 1013-

1022. 1982.

34. Tricot. G.. Orshoven. B. V.. Daele. C. V., and Van Den Berghe. H. 8:21 translocationin acute myeloid leukaemia. Scand. J. Haematol.. 26: 168-176. 1981.

35. Groupe Françaisde Cytogenetique Hematologique. Acute myelogenous leukemiawith an 8:21 translocation. Cancer Genet. Cytogenet., 44: 169-179, 1990.

36. Davey. D. D., Patii, S. R., Echternacht. H., Fatemi. C., and Dick. F. R. 8:21Iranslocation in acute nonlymphocytic leukemia. Am. J. Clin. Pathol.. 92: 172-176.

1989.37. Marlton, P., Keating. M.. Kantarjian. H.. Pierce. S., O'Brien. S., Freireich. E. J.. and

Estey, E. Cytogenetic and clinical correlates in AML patients with abnormalities ofchromosome 16. Leukemia (Baltimore). 9: 965-971. 1995.

38. Ghaddar. H. M.. Plunkett. W., Kantarjian, H. M., Pierce. S., Freireich. E. J.. Keating.M. J., and Estey, E. H. Long-term results following treatment of newly-diagnosedacute myelogenous leukemia with continuous-infusion high-dose l-ß-i>-arabino-furanosylcytosine. Leukemia (Baltimore). 8: 1269-1274. 1994.

39. Caligiuri, M. A., Strout, M. P., Lawrence, D., Arthur. D. C.. Baer. M. R., Yu. F.,Knuutila, S., Mrózek, K.. Oberkircher. A. R.. Marcucci, G., de la Chapelle. A.,Elonen. E., Block. A. W.. Rao, P. N., Herzig. G. P.. Powell. B. L.. Ruutu. T., Schiffer.C. A., and Bloomfield. C. D. Rearrangement of ALLÕ (MLLl in acute myeloidleukemia with normal cytogenetics. Cancer Res.. 58: 55-59. 1998.

4179

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from

1998;58:4173-4179. Cancer Res   Clara D. Bloomfield, David Lawrence, John C. Byrd, et al.   Cytogenetic SubtypeCytarabine Intensification in Acute Myeloid Leukemia Varies by Frequency of Prolonged Remission Duration after High-Dose

  Updated version

  http://cancerres.aacrjournals.org/content/58/18/4173

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/58/18/4173To request permission to re-use all or part of this article, use this link

Research. on January 2, 2021. © 1998 American Association for Cancercancerres.aacrjournals.org Downloaded from