From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.

1993 82: 398-407

Randomized comparison of busulfan and hydroxyurea in chronic myelogenous leukemia: prolongation of survival by hydroxyurea. The German CML Study Group
R Hehlmann, H Heimpel, J Hasford, HJ Kolb, H Pralle, DK Hossfeld, W Queisser, H Loffler, B Heinze and A Georgii

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.

Randomized Comparison of Busulfan and Hydroxyurea in Chronic Myelogenous Leukemia: Prolongation of Survival by Hydroxyurea
By R. Hehlmann, H. Heimpel, J. Hasford, H.J. Kolb, H. Pralle, D.K. Hossfeld, W . QueiBer, H. Lbffler, B. Heinze, A. Georgii, P.v. Wussow, C. Bartram, M. GrieBhammer, L. Bergmann, U. Essers, C. Falge, A. Hochhaus, U. QueiBer, C. Sick, P. Meyer, N. Schmitz, K. Verpoort, H. Eimermacher, F. Walther, M. Westerhausen, U.R. Kleeberg, A. Heilein, A. Kabisch. C. Barz, R. Zimmermann, G. Meuret, A. Tichelli, W.E. Berdel, L. Kanz, B. Anger, F.J. Tigges, L. Schmid, W. Brockhaus, R. Zankovich, U. Schlafer, 1. WeiBenfels, K. Mainzer, A. Tobler, M. Perker, J. Hohnloser, D. Messener, J. Thiele, T. Buhr, H. Ansari, and the German CML Study Group
In a randomized multicenter study the influence of hydroxyurea versus busulfan on the duration of the chronic phase and on survival of chronic myelogenous leukemia (CML) was determined. In addition cross resistance and adverse reactions of the drugs were analyzed. From July 1983 to January 1991, 441 CML patients were randomized to receive hydroxyureaor busulfan. Of these, 90.7% were Philadelphia positive; 25.7% were low, 38.2% intermediate, and 36.2% high risk patients according to Sokals score. The median survival of the busulfan treated Philadelphiapositive patients is 45.4 months and of the hydroxyurea group 58.2 months (P = .008). The survival advantage for the hydroxyurea treated patients is recognized in all risk groups. Sixty four patients reached therapy resistance before blast crisis and were crossed over to the alternative drug. The 23 patients with primary hydroxyurea had a median survival of 5.6 years, the 4 1 patients with primary busulfan therapy a median survival of 2.7 years (P = .02). Adverse reactions were less frequent with hydroxyurea with no severe adverse effects (lung fibrosis, long lasting bone marrow aplasia). The analysis of white blood cell counts in the course of treatment showed lower counts in the hydroxyurea patients. W e conclude that hydroxyurea is superior to busulfan in therapy of CML in chronic phase and should be used as first line therapy. Busulfan may have a role as secondarytherapy after hydroxyurea resistance or intolerance. 0 1993 by The American Society of Hematology.

RUG THERAPY of chronic myelogenous leukemia (CML) has basically been palliative until now, with the drug of choice being busulfan for the past 40 years. Little prolongation of life has been observed since the report by Minot et al in 1924. Curative therapy can be offered only to those 10%to 20% of CML patients of sufficiently young age, who have an HLA compatible, related or unrelated, bone marrow donor.z3 Median survival of CML patients from the time of diagnosis varied between 30 and 55 months, depending less on the mode of therapy than on patient selection and exclusion of high risk patients such as Philadelphia-negative or preblastic ones.4 Busulfan, an alkylating agent active at the stem cell level, is known to control CML-related signs and symptoms in 95% of patients for at least 3 months. All studies have shown its efficacy and reliability since its introduction in 1953. Controlled comparison with a number of single agents, radiotherapy, and with intensive combination therapy has shown its superiority with regard to efficacy, adverse reactions, and/or duration of disease control. The me-

From Klinikum Mannheim, Universitat Heidelberg, Mannheim, Deutschland. Submitted December 21, 1992; accepted February 25, 1993. The study was initiated by the Sueddeutsche Haemoblastosegruppe and is supported by the German Bundesminister fuer Forschung und Technologie, Foerderkennzeichen 01Z W044, 01ZP900I -01ZP90019, and 01Z W8503. Address reprint requests to R. Hehlmann, MD, III. Medizinische Klinik, Klinikum Mannheim der Universitiit Heidelberg, Wiesbadenerstr. 7-11, 6800 Mannheim, Germany. The publication COSIS of this article were defiayed in part by page charge payment. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 I993 by The American Society of Hematology. 0006-4971/93/8202-0031$3.00/0
398

dian (or mean) survival times of busulfan-treated CML patients range from 35 to 47 months4 More recently, hydroxyurea, an inhibitor ofthe ribonucleotide reductase, became increasingly popular because of its rapid action and low level of adverse effects. Some retrospective studies on small series of patients indicated a survival advantage of hydroxyurea-treated patients,12- which, however, was never substantiated by a controlled study. As early as 1972, Kennedy reported in a retrospective study on 20 patients with unknown Philadelphia chromosome status that hydroxyurea controlled all CML-related symptoms as well as busulfan. Hydroxyurea did not prevent progression to blast crisis, but the investigator felt that it could prolong the duration of the chronic phase as compared with busulfan. The median duration of response to hydroxyurea was 41 months for 10 not pretreated patients and 8 months for 10 pretreated and busulfan-resistant patients. Schwarzenberg et al, in a study of 43 patients who were treated with hydroxyurea, leukapheresis, and splenectomy, reported that survival was similar to that after busulfan therapy,l3 Schwarz and Canellos14treated 35 patients, mostly resistant to busulfan and/or in accelerated phase, with hydroxyurea and had good results. Bolin et all5 reported the results of a retrospective study on 30 busulfan- and 14 hydroxyureatreated patients. Expected median survival was 48 months and >90 months for the busulfan and the hydroxyurea groups, respectively. However, this difference was not significant due to the small sample size. In 1983, the German CML study group, therefore, decided to compare in a randomized study the influence of hydroxyurea versus busulfan on the duration of the chronic phase and on survival of CML. Further goals of the study were the examination of cross resistance of hydroxyurea and busulfan and the determination of duration of efficacy after cross over; the development of a prognostic score on the basis of prospectively documented parameters of possible significance for the prognosis of CML; the comparison
Blood, VOI 82,

NO 2 (July 15). 1993:pp 398-407

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
CML: PROLONGATION OF SURVIVAL BY HYDROXYUREA

399 CML patients had exclusion criteria recognized after randomization: not in chronic phase (acceleratedor blastic phase, 9 patients), decision of patient (1 patient), and other reasons ( 1 patient). One hundred seven patients were censored (52 in the hydroxyurea arm, 55 in the busulfan arm) when the followingevents happened bone marrow transplantation (44 patients), diagnosis of second neoplasia, withdrawal of consent, or noncompliance (63 patients). One hundred ninety two patients died- 13 I in blast crisis, 34 of CML independent causes. Other causes of death (20 patients) were infection, hemorrhage, marasm, bone marrow aplasia, and bone marrow fibrosis. In seven patients the cause of death was unknown. One hundred forty two patients are in-study and alive (88 in the hydroxyurea arm, 54 in the busulfan arm). The present report is based on the 44 1 documented and randomized CML patients on an intention-to-treat basis. The median observation time of all 441 patients is 2.03 years (range 0-7.83 years). Diagnostic investigations. Pretherapeutic diagnosticevaluation consisted of history, physical examination, complete blood count including reticulocytes, alkaline leukocyte phosphatase (ALP), and lactate dehydrogenase (LDH), chromosome analysis, and bone marrow cytology and histology. Review panels controlled quality of bone marrow histology and chromosome analyses. Follow up investigationswere performed and documented every 6 months, at 12 months, and at disease progression (checkpoints I , 2, and 3). Investigations at 6 month intervals included inquiry for symptoms (fever, fatigue, drug side effects), physical examination (spleen size, extramedullary manifestations,adverse drug reactions, weight), complete blood count including reticulocytes, and LDH. Investigationsat 12 month intervals included chromosome analysis and bone marrow cytology and histology. In addition, the annual dosages of the respective drugs were determined. Therapy. Busulfan was administered in a dosage of 0. I mg/kg/ d. The dosage was reduced when thrombocytopenia (<IO0 X IO9 platelets/L) was present. The initial dosage was continued until the WBC count was reduced by 50%, and then it was reduced by 50%. WBC counts were controlled weekly. The dosage was further reduced by 50%with each further reduction of WBC counts by 50%. Therapy was discontinued, when the WBC count dropped below 20 X 109/L.Therapy was renewed at a WBC count of more than 50 X 109/L. If there was no sufficient therapeutic response or disease progression the dosage was increased up to twice the initial dose. In

of the terminal phases of CML under the different treatment modalities; the determination of adverse drug reactions; and the comparison of the outcome of bone marrow transplantations after the different drug therapies. In the present report the results of the comparison of hydroxyurea with busulfan, and in particular, the prolongation of survival of CML patients treated with hydroxyurea will be reported.
PATIENTS AND METHODS

Study design. The study outline is shown in Fig 1 (see Hehlmann et all6 for comparison). After development of resistance to the randomized drug, therapy was crossed over to the other drug. Checkpoints were 1) the end of the chronic phase, defined as resistance to doubling of the initial dose of the randomized drug with cross over of drugs (hydroxyurea for busulfan resistant patients, busulfan for hydroxyurea resistant patients); 2) resistance to the alternative therapy; 3) diagnosis of blast crisis; and 4) death. Blast crisis was diagnosed, if blasts and promyelocytes were more than 30%of peripheral white blood cells (WBCs), or more than 50% of nucleated cells in the bone marrow. Patients with bone marrow transplantations were censored as lost to follow up at the time of transplantation. Patients. All newly diagnosed patients with CML in chronic phase were randomized to receive either busulfan or hydroxyurea when they fulfilled at least one of the following six criteria: general ill feeling and fatigue, weight loss of more than 10% in 6 months, unexplained fever of more than 38.5"C on 5 consecutive days, organomegaly related symptoms, leukocytosisof more than 50 X IO9/ L, thrombocytosis of more than 1 X IO'*/L. Exclusion criteria were no chronic phase (n = 5), no treatment required (n = IO), pretreatment with cytostatics, interferon alpha or splenic irradiation (n = 6), lack of informed consent (n = 38), second neoplasia (n = IO), and other reasons that made a therapy according to protocol a priori unlikely (n = 1 I). From July 1983 to January 1991,458 patients were randomized by 60 centers in Germany and Switzerland, 226 to receive busulfan and 232 hydroxyurea (see flow diagram in Fig 2). A third arm with interferon alpha was opened later and is not yet evaluable. Eleven

R
A N D 0 MJ I
A
~

Checkpoint 1 Busulfan

Checkpoint 2

Checkpoint 3

Checkpoint 4

When treatment required

S L Hydroxyurea
0

r r Therapy I I )Therapy I I )Diagnosis resistance resistance of blast I against crisis alternative

+Death

T I
Fig 1. Study outline.

Cross over of drugs

""i
Free therapy

BMT-patients counted as "lost to follow up"

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
400 HEHLMANN ET AL

Total no. of randomized patients 622

I F N a
164

Hydroxyurea

Busulfan
(sv)
226

(HU)
232

Evaluable patients

\/
Patients entered in the intcntionto-treat analysis
HU BU

I 1

Excluded because of rejection of CMLDiagnosis

Hu
5

BU

0.05. The required 01 for the final analysis was 0.04009. The analyses followed the intention-to-treat strategy. Survival was analyzed with the Kaplan-Meier estimator and logrank test.20Analyses for prognostic factors were performed with Cox proportional hazards regression model.2 The calculation of risk groups was performed with the equation? Score = exp. [0.0116 (age, 43.4) + 0.0345 (spleen size, 7.51) + 0.0188 ((platelet~/700)~ - 0.563) + 0.0887 (peripheral blasts, 2. IO)]. All analyses were performed by the Biometric Center with the program package SAS. Ethics. The protocol followed the declaration of Helsinki and was approved by the ethics committees of the Universities of Munich and Ulm. Informed consent was obtained from all patients.
RESULTS

21

216

Drop&t (censored)
HU
52

Death (event)
HU

Alive (censored)
Hu
Q

BU 55

BU
107

BU
54

85

Initial patients characteristics. The initial characteristics of the 44 1 randomized and documented CML patients are shown in Table 1 according to treatment group and Philadelphia (Ph) status. The Ph status was known for 409 patients of which 37 1 (90.7%) were Ph positive ( 1 87 in the hydroxyurea, 184 in the busulfan arm). All patients are shown including the nine patients who were not in chronic

4
Death after dropout
Hu
12

Table 1. C M L Initial Patients Characteristics

BU (n = 216)
HU (n = 225)

BU 8

Phf (n = 371)

Fig 2. Flow diagram of randomized patients.

the presence ofrapidly rising WBC counts (cell doublingin less than 2 months) low dose continuous busulfan therapy was acceptable. Hydroxyurea was administered at an initial dosage of 40 mg/kg/ d. The initial dosage was reduced in the presence of thrombocytopenia (< 100 X IO9 platelets/L). Control ofthe blood counts initially was three times weekly due to the rapid action of hydroxyurea. When the WBC count dropped to below 20 X 109/L the hydroxyurea dosage was reduced to 20 mg/kg/d. This dosage was then adapted individually. If WBC counts rose above 20 X 109/L, the dosage was increased. The dosage was reduced, if the WBC count dropped below IO X 109/L, and was interrupted, if the WBC count was below 5 X IO/L. Therapeutic goal was a normal WBC count (range 5 to 15 x lo9&). Documentation, At diagnosis and randomization all data were documented in an initial documentation form. During the course of the study documentation forms were completed every 6 months. Whenever a checkpoint was reached (see study outline) a separate checkpoint form was filed. After a patient had died or was lost to follow up (including bone marrow transplantation), a final documentation form was filed. Adverse events were documented according to WHO grading. Biostatistics. Sample size estimation was done according to George and Desu. Assuming 01 = 0.05 (two-sided) and 6 = 0.20, 388 patients (194 per group) were necessary to detect a relative risk of at least 1.42 in the median survival time in favor of hydroxyurea. The randomization lists were computed according to Efronl stratified for participating hospitals. After getting informed consent suitable patients were randomized centrally by phone by the Biometric Center for Therapeutic Studies. Five interim analyses were performed as planned in the trial protocol. The probability for type I error 01 was adjusted following OBrien and Fleming, safeguarding an overall error probability <

Age (vr) Sex (% male) Fatigue, general ill feeling (%) Symptoms due to organomegaly (%) Weight loss (%) Fever (%) Karnofsky index (%) Splenomegaly (%) Hepatomegaly (%) Spleen size in cm below costal margin Liver size in cm in MCL Extramedullary manifestations (skin, lymphnodes)(%) Ph-positive (%)* Additional cytogenetic aberrations (%) WBC x IO=/L Platelets x 109/L Hemoglobin (g/dL) LDH (U/L) Circulating blasts (%) Circulating promyelocytes (%) Basophils (%) Reticulocytes (%o) Erythroblasts (per 100 WBC) ALP (index) BM blasts (%)t BM promyelocytes (%)t Grouping into risk groups (Ph+ only)* Low (%) Intermediate (%) High (%I

50.2 60.7 69 37 24 9.4 87.1 72.2 54.6 6.4 12.9 9 91.5 11 161.9 454.3 12.2 725 2.9 4.5 4.5 23.0
1.1 15.2 4.1 10.1

49.2 52 59 34.7 20.7 4.5 88.7 72.4 46.6 6.3 12.6 3.6 89.9 12.9 156.6 488.4 11.9 720 3.4 4.5 4.5 23.7 0.9 16.0 4.4 10.7 26.0 35.0 39.0

47.8 58.8 64 36 19 7 88.2 71.9 49.2 6.5 12.6 6 100 9.8 169.1 493.9 12.1 749 3.1 4.7 4.9 23.3 1 .o 15.5 3.9
10.1

25.3 41.6 33.1

25.7 38.2 36.2

All features not specified as otherwise were recorded in 95% to 100% of patients. Features recorded in 93% (cytogenetics)and 80%(ALP) of patients. t Features recorded in 74% of cases. t Features recorded in 92.5% of Ph+ patients.

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
CML: PROLONGATION OF SURVIVAL BY HYDROXYUREA

40 1

n = 343
p < 0.0001

"?

> 1,2, n = 124

Fig3. Characteristicsof Philadelphia-positive CML patients according to Sokal's prognostic subgrouping.

0.0
0

,,,,,,,,,

, , , , , I, , , , , , , , , , , , , , , , , , , ,,,,,,,,,

I,

,,,,,,,,,,,,,,,,,,

,,,,I,.,,.,
7

a
v

phase at randomization. The patients' characteristics are distributed evenly in the two randomization groups. Figure 3 shows the characteristics of the Ph-positive patients according to Sokal's prognostic subgrouping." The proportion of low risk patients is 25.7%.The distribution of busulfan and hydroxyurea treated patients is similar in all three risk groups (Table 1). There are somewhat more hydroxyurea patients in the high risk group (39% vs 33%). Survival. The median survival of all CML patients is 43.6 months in the busulfan group and 56.2 months in the hydroxyurea group (P = .Ol); that of the Ph-positive patients in the busulfan group is 45.4 months and in the hydroxyurea group 58.2 months (P = .008, Fig 4). Duration of the chronic phase. Figure 5 shows the duration of the chronic phase defined by resistance to the ran1 . 0 -. II+~~WIH

domized therapy or diagnosis of blast crisis (checkpoint 1,2, or 3). The median duration from diagnosis to checkpoint 1, 2, or 3 is 37 months for the busulfan and 47 months for the hydroxyurea group (P = .04). Analysis according to risk factors. The survival advantage in the hydroxyurea arm is recognized in all Sokal prognostic subgroups (Fig 6). In spite of the smaller numbers in the subgroups, the survival difference in the intermediate risk group is significant at P = .O1. Checkpoint I and crossover o f therapy. Sixty four patients reached checkpoint l first, 41 in the busulfan and 23 in the hydroxyurea group, and were crossed over to the other therapy. The mean duration of therapy before crossover was 2 1.1 months in the busulfan and 22.3 months in the hydroxyurea group. The mean duration of secondary

- Hydroxyurea, n

= 187, median: 58.2 months

0.9:
0.8:

..___ Busulfan, n = 184, median: 45.4 months

p = 0.008

0.7:
0.G-

0.5:

0.40.3:
0.2:

'!$.,; ....

L.

.......
,,,,,,
,,,,,,,,,I ,,,,,,,,, ,,,,, ,,,, ,,,,,,,,
I I

0.I-.

...4

0.0 0

1 , ,

,,,(,,,,,,,,,,I , , , I I ,,,,

7
Years

Fig 4. Survival of Philadelphia-positive CML patients according to treatment group.

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
402
HEHLMANN ET AL

Hydroxyurea, n = 225, median: 47 months

............ Busulfan, n = 216, median: 37 months

p = 0.04

0.2:
0.1:

:.......,..........

. . . . .

.,,...........,
8
Fig 5. Duration of chronic phase (time from diagnosis to resistance to randomized therapy or blast crisis).

years

busulfan therapy was 20 months, that of secondary hydroxyurea therapy I I .7 months. Eleven patients progressed to blast crisis within 1 month after crossover (four in the group with primary hydroxyurea therapy, seven in the group with primary busulfan therapy). Because of secondary resistance before blast crisis (checkpoint 2) interferon alpha was added in seven cases (four busulfan patients, three hydroxyurea patients), other chemotherapy (arabinosylcytosin, mitomycin, thioguanin) in 10 patients (five in each therapy group). It is evident from Fig 7 that secondary busulfan after hydroxyurea resistance has an additional impact on survival. There is a survival advantage for primary hydroxyurea and secondary busulfan in all subgroups of Table 2. Most notably, patients with cytopenia and/or bone marrow aplasia, who carry a poor prognosis, are only found in the primary busulfan group. But even without the 16 cytopenic patients, the survival difference between the treatment groups after crossover is significant (P = .01). Once blast crisis has developed no survival difference is observed between the treatment groups (data not shown). Cytogenetic response rates. Four patients in the hydroxyurea arm and two patients in the busulfan arm showed cytogenetic responses (reduction of Ph-positive cells by 10% to 100%)during the course oftherapy corresponding to 3% and 2%of the evaluable patients. One complete cytogenetic remission was observed in a hydroxyurea treated case. Some of the responses may be due to additional therapy added to the monotherapy later on when secondary resistance was observed (checkpoint 2) (the one case with a complete cytogenetic remission also had polychemotherapy, another case also had interferon alpha). Adverse effects. Data were available from 209 patients treated with hydroxyurea and from 204 patients treated with busulfan. The frequency of adverse effects is lower with hydroxyurea than with busulfan ( 1 5.8 vs 24.2 symptoms per 100 patient years). Most importantly, serious adverse effects such as long lasting bone marrow aplasia or lung fibrosis were virtually not observed in the hydroxyurea arm (one

transient event under hydroxyurea vs 13 events in the busulfan arm). This low toxicity may have facilitated a possibly more intensive treatment with hydroxyurea. Correlation of survival with WBC counts. Since one reason for the difference of survival might be a difference in disease control, we analyzed the WBC counts in both groups during the first 24 months of treatment. After 18 and 24 months 43%and 34%of the hydroxyurea-treated, but only 11% and 16% of the busulfan-treated patients had normal WBC counts, respectively. In the hydroxyurea group patients with normal WBC counts had a survival advantage compared with those with elevated counts (P< .06).
DISCUSSION

The most important observation in this study is the significantly better median survival of the patients treated with hydroxyurea as compared with those treated with busulfan. This result confirms some reports of a possible survival advantage by hydroxyurea in small, uncontrolled but the magnitude of the survival advantage came as a surprise. The result raises several biologic as well as therapeutic questions. First, the question has to be addressed whether hydroxyurea prolongs the survival of CML patients or whether busulfan therapy was suboptimal. The median survival time of 45.4 months with busulfan compares well with an earlier controlled study that finds a median survival of busulfantreated CML patients of 44 months.6 In addition, it has to be excluded that the survival difference is due to an unequal distribution of prognostic factors. In Table 1 we show that the relevant prognostic parameters as well as the distribution of risk groups did not favor the hydroxyurea group. There were more hydroxyurea patients in the high risk group (39%vs 33%),although this was not significant. We therefore conclude that the survival difference is the result of the different therapies. Second, the apparent prolongation of survival of the hydroxyurea group is contrary to some earlier notions on pro-

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
CML: PROLONGATION OF SURVIVAL BY HYDROXYUREA

403

'.O
0.9

7;:;
L . . + . . + , .
~

. . . . . . . . . Busulfan, n = 42

:.7

. . . . . . . . . . . . .+-

m .2 e

0.8
0.1

i-4:

. . . .

\: . ....I..

0.6
0.5
0.4

.._____ .._.
* t :

1 :

> ..-

I.. . .

. . . . . . . . . . . . . . . I . . .

.
L.1

. . . . . . . . . . . . . . . . . . . . . . . .

L)

r)

0.3: 0.2:

. . . . . . . . . . . ,

I.o

0.9

0.8

> .-

0 7

0 6
0.5
0.4
!

=
D

2 0
x ..D

....

0.3
0.2
0. I
0.0

. . .
5 6
7

le
0
1

1.0

Hydroxyurea, n = 69 Busulfan, n = 55

C O

0 71

> .0.6-

tz 0
2 a
Fig 6. Survival curves of the treatment groups according to risk group (Sokal's prognostic score: [A] low risk; [B] intermediate risk; [C] high risk).

0.5:

. 2 0.4O.?..........
,,,,,

0.0

'-,
0

. , ,,
I

, , ,,.

... , . . . , . . . .
2

I " " " ' ' ' I " ' ' ~ ~ " ' ~ ~ " ' " ' " ~ ~ ~ ' " ~ ~ '

'
7

'

'

,
8

>

years

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
404
HEHLMANN ET AL

b1
0.8:

primary hydroyurea (followed by busulfan) n = 23. median survival 5.6 years

0.1:
0.60.5:0.4:
0 3....

p= 0.02

.........
...................
i lj
.(...,

............
............I + ................

,I,,,

, , , , , , , , , I , , , a , :, , , , , . , , , , , , , I , , , , , , , , ,

, , , , , ( , I ,

, , , , , , , ~, ,,,,,,,,,

years

- primary hydroxyurea (n = 23). median survival 2.45 years

___ primary . _ busulfan (n

= 41), median survival 0.9years

p = 0.002

0.6-

0.5:
0.40.3:

0.2-

........................

.__,
LI

.............................
I ,

I. . . . . I

I.

'

I -

* *

r r

"

'

1 ,

"

''

'

'

I rl-rm-mq-

7
years

Fig 7. Survival after crossover of therapy (checkpoint 1). (A) Survival from diagnosis; (B) survival from crossover.

gression of CML that the course of CML is determined by This has to be anaintrinsic factors rather than lyzed also in the context ofthe good survival data with interOne interferon alpha (our unpublished pretation would be that a decrease of tumor mass and/or a slowing of granulocyte proliferation rate decreases the progression rate to blast crisis. The hypothesis would be that the number of clonal cells in the mitotic pool correlates with likelihood of blastic transformation implying that the less tumor burden and proliferation are present the less likely blastic transformation might occur. This hypothesis is supported by the observation of Kolitz et a129 that intensive hydroxyurea therapy induces cytogenetic responses in nine of 14 patients of 25% to 100%and that the level of bone marrow hypocellularity correlates with cytogenetic response. If this hypothesis is true, the more efficient reduction of WBC counts, which probably reflect tumor cell mass, by hydroxyurea might in part account for the survival advantage.

Therefore, if there is a correlation between tumor mass and probability of blastic transformation, a more efficient reduction of WBC counts by the combination of suitable antiproliferative drugs might result in a further prolongation of survival in CML (eg, hydroxyurea and interferon alpha as in our ongoing randomized CML study I1 or combinations with arabinosylcyt~sine).~~ In this context, the question has to be addressed why intensive combination chemotherapy did not show prolongation of survival as observed with h y d r o x y ~ r e a . ~ The ~'~~~'~~~ goal of these studies, however, primarily was cure of CML by the intensive approach and not continuous control of myeloproliferation over extended periods of time as in our study. Finally, the high mutagenic potential of busulfan with subsequent promotion of acute leukemia as observed in polycythemia verag3has to be considered. Although the median survival time of the busulfan-treated patients of more than 45 months is longer than that of historic controls of

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
CML: PROLONGATION OF SURVIVAL BY HYDROXYUREA

405

Table 2. Reasons for, and Survival After, Crossover of Therapy Before Blast Crisis
Patients (mean survival after crossover in years) Primary Busulfan Primary Hydroxyurea

Reasons for Crossover

C
18

Progressive thrombocytosis' Resistance with progression of disease Resistance with progression of disease and cytopenia and/or bone marrow aplasia Various reasonst

lO(1.8) 9 (0.4)

8(2.6) 9 (2.1)

therapy of CML in chronic phase and should be used as first line therapy. Busulfan may, however, have a role as second line treatment after hydroxyurea resistance or intolerance. The importance of the reduction of WBC in the chronic phase for prolongation of survival in CML, irrespective of the cytostatic drug used, will be analyzed further within ongoing randomized studies.
ACKNOWLEDGMENT

18

16 (0.7) 6(1.7) 41 (1.0)

6(2.4) 23 (2.2)

16 12 64

The assistance of Dr A. Kiittel, B. Knichel, M. Dumke, and G. Alter is gratefully acknowledged.

APPENDIX

Platelets of in general more than 1 million/pL or rapidly rising platelets above ca. 700.000/pL in spite of intensificationof therapy (doubling of initial dose). t Intolerable adverse reactions ( 5 ) . primary (a priori) drug resistance (3). decision of patients (2),BMT ( I ) , unknown (1).

untreated CML patients,' or of CML patients treated with splenic irradiation,'.' it cannot be excluded that its mutagenic effect has a negative impact on survival, which becomes visible, if busulfan is compared with hydroxyurea. A third observation of interest is the apparent gain of survival time by secondary busulfan after primary hydroxyurea therapy. Although the number of patients qualifying for an adequate trial with secondary busulfan is small (n = 19, 8.4% of hydroxyurea-treated patients), the survival advantage is significant and adds to the survival advantage of the hydroxyurea arm described in this report. Busulfan turned out to be inferior to hydroxyurea as first line therapy, but apparently plays a role after hydroxyurea resistance. One reason for the poor success of secondary hydroxyurea treatment after primary busulfan seems to be the high number of patients with cytopenias and/or bone marrow aplasia after busulfan therapy, which was not observed after primary hydroxyurea therapy. This agrees well with the known toxicity pattern of busulfan.24It appears that in the presence of busulfan-induced cytopenias hydroxyurea cannot add any survival advantage. We cannot entirely exclude that our survival results are influenced by an earlier diagnosis than in previous eras. Our inclusion criteria for randomization and therapy, however, excluded patients who were very early in the course of disease. In addition, the percentage of our patients in the lowrisk group is in contrast to Sokal's patient^'^.^^ lower than that in the intermediate- and high-risk groups. This distribution argues against the possibility that our favorable survival results are due to a high number of low-risk patients or earlier diagnosis. Since all parameters included in this report have been documented prospectively under the same mandatory modalities, the patient sample is well suited for the generation of a prognostic score (score 1) based on prospectively collected parameters35 as well as a detailed histomorphologic characterization of CML according to Ph status and clinical course, which is ongoing. We conclude that hydroxyurea is superior to busulfan in

Participating institutions: Hamatolog.-onkol. Praxis, Aachen; Abteilung fur Onkologie/Abteilung fur Hamatologie, Kantonsspital Basel; Onkologische Praxis Basel; Abt. Innere Medizin U. Poliklinik, Universitatsklinikum Rudolf Virchow, Berlin; Hamatologische Praxis, Berlin; Medizinische Klinik, St. Joseph Hospital, Bremerhaven; Med. Klinik 11, St. Johannes Hospital, Duisburg; Abt. Hamatologie/Onkologie, Med. Universitatsklinik, Freiburg; Zentrum der Inneren Medizin, Abt. f. Hamatologie U. Onkologie, Frankfurt; Med. Klinik IV mit Hamatologie und Onkologie, GieBen; Medizinische Univ.-Klinik, Hamburg-Eppendofi, Hamatolog.-onkol. Praxis Altona, Hamburg; Allgemeines Krankenhaus Barmbek, Hamburg; Onkologische Ambulanz, Med. Univ. Klinik, Heidelberg; Medizinische Poliklinik, Heidelberg; Innere Abteilung, Stadtkrankenhaus Kempten; Med. Klinik 11, Universitilt &el; I. Med. Klinik der Universitat Koln; 11. Med. Klinik, Klinikum Karlsruhe; 111. Medizinische Klinik, Klinikum Mannheim der Universitat Heidelberg, Mannheim; 111. Med. Klinik, Klinikum GroBhadern der Universitat Miinchen; Med. Klinik Innenstadt der Universitat Munchen; Med. Poliklinik der Universitat Miinchen; Miinchner Onkologische Praxis; Med. Klinik, St. Elisabethen Krankenhaus, Ravensburg; Onkologische Abteilung, Kantonsspital St. Gallen; Innere Medizin 111, Med. Klinik U. Poliklinik, Ulm; Med. Poliklinik der Universitat Wurzburg; Innere Abteilung, Kreiskrankenhaus Aalen; Onkolog. Gemeinschaftspraxis, Frankfurt/M.; Allgem. Krankenhaus Altona, Hamburg; Onkolog. hamatolog. Schwerpunktpraxis, Berlin; I. Med. Abteilung, Stildt. Krankenhaus Schwabing, Miinchen; IV. Med. Abt., Stildt. Krankenhaus Miinchen-Harlaching; St. Willehad Hospital, Wilhelmshaven; Hamatologische Praxis, Miinchen; I. Med. Klinik, Klinikum der Christian-Albrechts-Universitat Gel; IV. Medizin. Abt., Krankenhaus Miinchen-Neuperlach; Innere Abt., Kreiskrankenhaus Boblingen; 11. Innere Abteilung, Krankenhaus Berlin-Neukolln; Hamatologisches Zentrallabor der Universitiit, Inselspital Bern; V. Med. Klinik des Klinikums Niirnberg; Stadt. Krankenhaus Kaiserslautern, Med. Klinik D 6, Kaiserslautern; Zentrum Innere Medizin, Abt. f. Himostaseologie, Klinikum Niirnberg; Klinik f. Innere Medizin, Onkologie/Hamatologie, Stildt. Kliniken Oldenburg; Kinderklinik und Poliklinik, Universitats klinikum Rudolf Virchow, Berlin; Kreiskrankenhaus Fussen; Innere Abteilung, Luisenkrankenhaus, Lindenfels/

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
406 HEHLMANN ET AL

Odenwald; Poliklinik der Universitat Erlangen; Klinik f. Hamatologie/Onkologie, St. Antonius Hospital, Eschweiler; I. Med. Klinik, Abt. f. Hamatologie U. Onkologie, Klinikum rechts der Isar, Munchen; Onkologische Abteilung, Krankenhaus der Barmherzigen Briider, Regensburg; Innere Abteilung, Stadt. Krankenhaus, Penzberg; Medizinische Klinik B, Klinikum der Landeshauptstadt Wiesbaden; Abt. Innere Medizin 11, Marien Hospital, Hagen; Med. Klinik, Hans-Susemihl-Krankenhaus, Emden; Praxisgemeinschaft Dr. Zankovich, Koln; Innere Abteilung, Diakonie-Krankenhaus, Schwabisch Hall; Medicina interna ed ematologia FMH, Lugano; I. Medizinische Klinik, Zentralkrankenhaus St. Jurgen-StraBe, Bremen; Institut fur Med. Informationsverarbeitung, Statistik und Biomathematik der LMU und Institut fur Med. Statistik und Epidemiologie der TU uber Biometrisches Zentrum fur Therapiesstudien (BZT), Munchen; Abt. f. Klinische Physiologie U. Arbeitsmedizin der Universitat Ulm; Immunhamatologisches Labor der Med. Klinik und Poliklinik, Universitat Ulm; Pathologisches Institut der MHH, Hannover; Abt. f. Knochenmarksdiagnostik an der Med. Klinik Innenstadt der Universitat Munchen; Institut f. Klinische Immunologie der MHH, Hannover; Pathologisches Institut der Universitat Koln; Sektion Molekularbiologie/Padiatrie 11, Blutspendezentrale Ulm.
REFERENCES
I . Minot GR, Buckman TE, Isaacs R: Chronic myelogenous leukemia. Age, incidence, duration and benefit derived from treatment. JAMA 82:1489, 1924 2. Thomas ED, Clift RA: Indications for marrow transplantation in chronic myelogenous leukemia. Blood 732361, 1989 3. Beatty PG, Anasetti C, Hansen JA, Longton GM, Sanders JE, Martin PJ, Mickelson EM, Choo SY, Petersdorf EF, Pepe MS, Applebaum FR, Bearman SI, Buckner CD, Clift RA, Petersen FB, Singer J, Stewart PS, Storb RF, Sullivan KM, Tesler MC, Witherspoon RP, Thomas E D Marrow transplantation from unrelated donors for treatment of hematologic malignancies: Effects of mismatching for one HLA locus. Blood 8 1:249, 1993 4. Hehlmann R: Cytostatic therapy in chronic myelogenous leukemia: Review and perspectives, in Huhn D, Hellriegel KP, Niederle N (eds): Chronic Myelocytic Leukemia and Interferon. Berlin, Germany, Springer Verlag, 1988, p 102 5. Haut A, Abbott WS, Wintrobe MM, Cartwright GE: Busulfan in the treatment of chronic myelocytic leukemia: The effect of long term intermittent therapy. Blood 17: I , I96 1 6. Canellos GP, Young RC, Nieman PE, DeVita VT, Jr: Dibromomannitol in the treatment of chronic granulocytic leukemia: A prospective randomized comparison with busulfan. Blood 45: 197, 1975 7. Medical Research Councils Working Party for Therapeutic Trials in Leukemia: Chronic granulocytic leukemia: Comparison of radiotherapy and busulfan therapy. Br Med J (Clin Res) 1:201, 1968 8. Cunningham I, Gee T, Dowling M, Chaganti R, Bailey R, Hopfan S, Bowden L, Turnbull A, Knapper W, Clarkson B: Results oftreatment of Ph+ chronic myelogenous leukemia with an intensive treatment regimen (L-5 protocol). Blood 53:375, 1979 9. Goto T, Nishikori M, Arlin Z, Gee T, Kempin S, Burchenal J, Strife A, Wisniewski D, Lambek C, Little C, Jhanwar S, Chaganti R, Clarkson B: Growth characteristics of leukemic and normal he-

matopoietic cells in Ph+ chronic myelogenous leukemia and effects of intensive treatment. Blood 59:793, 1982 IO. Kantarjian HM, Velekoop L, McCredie KB, Keating MJ, Hester J, Smith T, Barlogie B, Trujillo J, Freireich EJ: Intensive combination chemotherapy (ROAPIO) and splenectomy in the management of chronic myelogenous leukemia. J Clin Oncol 3:192, 1985 1 I . Kennedy BJ: The evolution of hydroxyurea therapy in chronic myelogenous leukemia. Semin Oncol 19:2 I, 1992 (suppl9) 12. Kennedy BJ: Hydroxyurea therapy in chronic myelogenous leukemia. Cancer 29:1052, 1972 13. Schwarzenberg L, Math6 G, Pouillart P, Weiner R, Locour J, Genin J, Schneider M, de Vassal F, Hayat M, Amiel JL, Schlumberger JF, Jasmin C, Rosenfeld C: Hydroxyurea, leukapheresis and splenectomy in chronic myeloid leukemia at the problastic phase. Br Med J (Clin Res) 1:700, 1973 14. Schwartz JH, Canellos GP: Hydroxyurea in the management of the hematologic complications of chronic granulocytic leukemia. Blood 46:l I , 1975 15. Bolin RW, Robinson WA, Sutherland J, Hamman R F Busulfan Venus hydroxyurea in long-term therapy of chronic myelogenous leukemia. Cancer 50:1683, 1982 16. Hehlmann R, Anger B, Messerer D, Zankovich R, Bergmann L, Kolb HJ, Meyer P, Essers U, QueiDer U, Vaupel H, Walther F, Hossfeld DK, Zimmermann R, Heiss F, Mende S, Tigges FJ, Kleeberg UR, Pralle H, Kayser W, Tichelli A, Faulhaber JD, Rath U, Schubert H, Bross K, Schlag R, Schmid L, WeiGenfels I, Heinze B, Georgii A, QueiDer W, Heimpel H: Randomized study on the treatment of CML in chronic phase with busulfan versus hydroxyurea versus interferon alpha. Blut 56237, 1988 17. George SL, Desu MM: Planning of size and duration of a clinical trial studying the time to some critical event. J Chron Dis 27: 15, 1974 18. Efron B: Forcing a sequential experiment to be balanced. Biometrika 58:403, 197 1 19. OBrien PC, Fleming TR: Multiple testing procedure for clinical trials. Biometrics 35:549, 1979 20. Kalbfleisch JD, Prentice R L The Statistical Analysis of Failure Time Data. New York, NY, Wiley, 1980 2 I . Cox DR: Regression models and life-tables (with discussion). J Roy Stat SOC 34: 187, 1972 22. Sokal JE, Cox EB, Baccarani M, Tura S, Gomez CA, Robertson JE, Tso CY, Braun TJ, Clarkson BD, Cervantes F, Rozman C, the Italian Cooperative CML Study Group: Prognostic discrimination in good-risk chronic granulocytic leukemia. Blood 63:789, 1984 23. SAS Institute Inc: SAS/STAT Users Guide, version 6 , fourth ed, vol 1. Cary, NC, SAS Institute, 1989,943 24. Galton DAG: Chemotherapy of chronic myelocytic leukemia. Semin Hematol 6:323, 1969 25. Sokal JE: Evaluation of survival data for chronic myelocytic leukemia. Am J Hematol 1:493, 1976 26. Talpaz M, Kantarjian H, Kurzrock R, Trujillo JM, Gutterman JU: Interferon alpha produces sustained cytogenetic responses in chronic myelogenous leukemia. Ann Intern Med 114532, 1991 27. Alimena G, Morra E, Lazzarino M, Liberati AM, Montefusco E. Inverardi D, Bernasconi P, Mancini M, Donti E, Grignani F, Bernasconi C, Dianzani F, Mandelli F Interferon alpha-2b as therapy for Ph-positive chronic myelogenous leukemia: A study of 82 patients treated with intermittent or daily administration. Blood 72:642, 1988 28. Thaler J, Kiihr T, Gastl G, Huber H, Duba C, Kemmler G, Gattringer C, Fluckinger T, Niedenvieser D, Seewann H, Abbredens K, Lang A, Gadner H, Fereberger W, Schiller I, Hausmaninger H, Weitgasser R, Michlmayr G , Fridrik M, Huber C: Rekom-

From bloodjournal.hematologylibrary.org by guest on December 11, 2012. For personal use only.
CML: PROLONGATION OF SURVIVAL BY HYDROXYUREA

407

binantes Interferon a-2c bei Ph-positiver chronischer myeloischer Leukamie. Ergebnisse einer multizentrischen Phase-11-Studie. Deutsche Med Wochenschr 1 16:72 I , I99 1 29. Kolitz JE, Kempin SJ, Schluger A, Wong CY, Berman E, Jhanwar S, Arlin ZA, Gee T, Clarkson BD: A phase I1 pilot trial of high-dose hydroxyurea in chronic myelogenous leukemia. Semin Oncol 19:27, 1992 (suppl 9) 30. Guilhot F, Dreyfus B, Brizard A, Huret J-L, Tanzer J: Cytogenetic remissions in chronic myelogenous leukemia using interferon alpha-2a and hydroxyurea with or without low-dose cytosine arabinoside. Leuk Lymphoma 4:49, 1991 3 1. Baccarani M, Corbelli G, Tura S, and the Italian Cooperative Study Group on Chronic Myeloid Leukemia: Early splenectomy and polychemotherapy versus polychemotherapy alone in chronic myeloid leukemia. Leuk Res 5:149, 1981 32. Kantajian HM, Smith TL, McCredie KB, Keating MJ, Walters R, Talpaz M, Hester JP, Bligham G, Gehan E, Freireich EJ: Chronic myelogenousleukemia: A multivariateanalysisofthe asso-

ciations of patient characteristicsand therapy with survival. Blood 66:1326, 1985 33. Berk PD, Goldberg JD, Donovan PB, Fruchtman StM, Berlin NI, Wasserman L R Therapeutic recommendations in polycythemia vera based on polycythemia vera study group protocols. Semin Hematol 23:132, 1986 34. Sokal JE, Baccarani M, Tura S, Fiaccini M, Cervantes F, Rozman C, Gomez CA, Galton DAG, Canellos GP, Braun TJ, Clarkson BC, Carbonell F, Heimpel H, Extra JM, Fiere D, Nissen NI, Robertson JE, Cox E B Prognostic discrimination among younger patients with chronic granulocyticleukemia: Relevance to bone marrow transplantation. Blood 66: 1352, 1985 35. Hehlmann R, Heimpel H, Griesshammer M, Kolb HJ, Heinze B, Hossfeld DK, Wickramanayake PD, Essers U, Thiele J, Georgii A, Ansari H, Hochhaus A, Hasford J, and the German CML Study Group: Chronic myelogenous leukemia: Recent developments in prognostic evaluation and chemotherapy. Leukemia 6: I IOS, 1992 (suppl3)