acute leukemia and myelodysplasia in patients with a philadelphia chromosome negative chronic...
Click here to load reader
Post on 06-Jun-2016
Embed Size (px)
Acute Leukemia and Myelodysplasia in Patients With aPhiladelphia Chromosome Negative Chronic
Myeloproliferative Disorder Treated With HydroxyureaAlone or With Hydroxyurea After Busulphan
Iben Nielsen1 and Hans Carl Hasselbalch2*1Department of Hematology L, Rigshospitalet University Hospital, Copenhagen, Denmark
2Department of Medicine, Division of Hematology and Oncology, Roskilde University Hospital, Roskilde, Denmark
Eighty-three patients with various chronic myeloproliferative disorders [polycythemiavera (PV), essential thrombocytosis (ET), idiopathic myelofibrosis (IMF)] were analyzedfor the occurrence of acute myeloid leukemia (AML) and myelodysplasia (MDS) duringtreatment with hydroxyurea (HU) alone or HU following treatment with busulphan (BU). Atotal of 58 patients (29 PV, 14 ET, 12 IMF, 3 unclassified) had been treated with HU.Thirty-five of these patients had been treated with HU alone whereas 18 patients hadreceived both HU and BU. The follow-up period was 7.8 years. Twenty-five patients hadnot been treated with HU. In this patient group, 4 patients had been treated with BU. Thefollow-up period was 10.5 years. In the HU-treated group (n = 58) 7 patients developedAML and 5 patients MDS. Five of the 12 patients had been treated with HU alone, and 4patients had received both HU and BU. In the non-HU-treated group (n = 25) 1 patient withPV developed acute myeloid leukemia (AML). This patient had only been treated withphlebotomies. It is concluded that treatment with HU is leukemogenic, with an incidenceof AML and MDS of approximately 14% when used alone. The incidence is markedlyincreased to about 30% when HU is preceded by treatment with BU. HU is not recom-mended for use in younger patients, in whom non-leukemogenic agents such as -inter-feron and anagrelide should be used instead. Am. J. Hematol. 74:2631, 2003. 2003 Wiley-Liss, Inc.
Key words: chronic myeloproliferative disorders; hydroxyurea; acute myeloid leukemia;myelodysplasia
INTRODUCTIONSince the introduction of hydroxyurea (HU) for the
treatment of chronic Ph chromosome negative (Ph) my-eloproliferative disorders it has been debated whetherthis non-alkylating agent was leukemogenic or whether itwas a safe drug that was not associated with the devel-opment of myelodysplasia (MDS) or acute myeloid leu-kemia (AML) . In particular, several larger studiesin recent years have questioned whether HU is an inno-cent drug. Thus, it has now been established that HUadministered after or before treatment with alkylatingagents implies a considerable risk for later leukemictransformation [33,40]. Furthermore, a recent study hasfound that HU therapy may be associated with the de-velopment of a cytogenetic abnormality (17q-deletion), which further supports the idea that HU may indeedbe leukemogenic.
In many institutions, the potential leukemogenicity ofHU has influenced the choice of therapy, particularly inyounger patients in whom -interferon and anagrelideare considered to be safe and non-leukemogenic alterna-tives . However, as HU is still widely used for thetreatment of chronic Ph myeloproliferative disorders,also in younger patients, and the drug has been intro-duced for the treatment of sickle cell anemia and in thal-assemia-associated anemia, it is of utmost importance
*Correspondence to: Hans Carl Hasselbalch, Department of Medicine,Division of Hematology and Oncology, Roskilde University Hospital,Roskilde, Denmark. E-mail: [email protected]
Received for publication 28 October 2002; Accepted 15 April 2003
Published online in Wiley InterScience (www.interscience.wiley.com).DOI: 10.1002/ajh.10375
American Journal of Hematology 74:2631 (2003)
2003 Wiley-Liss, Inc.
that all cases of acute leukemia and MDS that are diag-nosed during treatment with HU are reported rapidly[35,41].
The objectives of the present study are (i) to report anadditional 12 patients, of a series of 58 patients withvarious chronic myeloproliferative disorders, who devel-oped AML or MDS after being treated with HU eitheralone or in combination with prior or subsequent busul-phan, (ii) to recommend not to use HU in younger pa-tients as the drug, even when given alone, appears to beleukemogenic, and (iii) to confirm previous observationsthat combination therapy with busulphan is highly leu-kemogenic and should be avoided.
MATERIALS AND METHODSFiles of all patients with chronic Ph myeloprolifera-
tive disorders referred consecutively to the Departmentof Hematology L, Rigshospitalet, from February 1993 toDecember 2000 were retrieved and analyzed retrospec-tively. All relevant clinical dataage at diagnosis, gen-der, administered cytotoxic treatment [only HU, busul-phan (BU), or -interferon (IFN) was given], andduration of the treatmentwere registered together withthe time interval between starting treatment and the de-velopment of AML or MDS.
The diagnoses of PV, ET, and IMF were made accord-ing to conventionally accepted international criteria . The diagnoses of MDS and AML were made ac-cording to French-American-British (FAB) criteria .
Patients were categorized into three separate groups:group 1, which had not received HU, group 2, which hadonly received treatment with HU, and group 3, whichencompassed those patients who had been treated withboth HU and BU.
PatientsThe study comprised a total of 83 patients. Fifty-eight
of the patients had been treated with HU. Twenty-fivepatients had never received treatment with HU. In theHU group, 36 patients were female (62%) and 22 weremale (38%). In the group of patients that did not receiveHU, 8 were female (32%) and 17 were male (68%).
In the group of patients being treated with HU, 29patients had a diagnosis of PV, 14 patients ET, 12 pa-tients IMF, and 3 patients had an unclassified myelo-proliferative disorder (Table I). In the control group, 1patient had ET, 13 patients had PV, and 11 patients hadIMF.
The average age at the time of diagnosis was virtuallyequal in the 2 groups of patients (64.4 years in the HU-treated group and 62.3 years in the control group).
The follow-up period in the control group was on av-erage 10.48 years (range 218 years). In the group ofHU-treated patients the follow-up periodas defined by
the time from starting HU treatment until the year2000was on average 7.8 years (range 27166 mo).Treatment
Of the 25 patients who never had been treated withHU, 13 patients had received treatment with phlebotomy,which in 3 patients was supplemented by treatment withBU. Four patients had only received treatment with bloodtransfusion, 1 patient had been splenectomized, and 4patients were not treated at all. Two patients were beingtreated with -interferon and recombinant human eryth-ropoietin. One patient had been treated with BU only.
Among the HU-treated patients, 35 had received HUas the only therapy. Sixteen patients had received BU incombination with HU, 1 patient had been treated withHU and anagrelide, 1 patient with HU and phlebotomy,1 patient with HU, BU, and IFN, 2 patients with HU,venesection, and BU, 1 patient with HU, venesection,and -IFN, and 1 patient with BU, HU, and anagrelide.
On average, each patient had received treatment with650.1 g HU (range 13,200 g). The average dosage ofHU in IMF patients was 576.4 g (range 621,775), in ET692.5 g (range > 352,000 g), in PV 701.3 g (range53,200 g), and in patients with an unclassified chronicmyeloproliferative disorder 252.2 g (range 1730 g).Statistical Calculations
All data were analyzed by the 2 distribution method.A P value
compared to the group of patients who had not receivedHU (P 0.033).
Out of the 12 patients who developed AML or MDS,5 patients had received HU as the only cytoreductivetreatment. Four patients had received HU in combinationwith BU, 1 patient had received HU in combination withIFN, 1 patient had received HU, BU, and IFN, whereas 1patient had been treated with HU in combination withBU and anagrelide. Accordingly, the incidence of AMLor MDS for patients receiving treatment with HU only orHU and phlebotomy was 13.8% (5/36), whereas the in-cidence of AML/MDS among patients, who had beentreated with several potentially leukemogenic agents(HU + BU; n 20) was 30%. No significant differencein the risk of developing AML was found between thosepatients treated with only HU and the group of patientswho, in addition to receiving HU, had received othercytoreductive treatment (BU; P 0.08). When thegroup that received no treatment is compared to thegroup that received both HU and BU, there is a signifi-cant difference in the risk of developing AML or MDS,as 1/25 patients in the non-treated group developed AMLas compared to 6/20 of the patients who received bothHU and BU (P 0.013).
On average the patients who developed AML or MDShad received 1,111 g HU (range 623,200 g), which wasvirtually twice the average for the whole group of pa-tients who had received treatment with HU (650.1 g).
AML or MDS developed on average 53.6 months aftertreatment with HU was instituted; the range was verybroad: from 5 months in a patient who developed AMLto 125 months in a patient who developed MDS. In 11 ofthe 12 patients, at least 19 months elapsed from startingHU treatment to development of AML or MDS.
On average, patients treated with HU (n 55) lived103 months from the time of diagnosis to either death orend of follow-up (range 4420 months), whereas patientsnot treated with HU (n 22) on average lived 73 months(range 8144 months) (P 0.05).This difference is inpart explained by 4 patients in the HU-treated group who
lived >20 years with a diagnosis of a chronic myelopro-liferative disorder. The median survival was equal in thetwo groups (72 months). Thus, patients treated with HUdid not live longer than those