A Systematic Overview of Chemotherapy Effects in

SYSTEMATIC OVERVIEW ARTICLE
A Systematic Overview of Chemotherapy Effects in

Acute Myeloid Leukaemia
Eva Kimby, Peter Nygren and Bengt Glimelius for the SBU-group1
From the Department of Haematology, University Hospital, Huddinge (E. Kimby), and Department of
Oncology, Radiology and Clinical Immunology, University Hospital, Uppsala (P. Nygren, B. Glimelius),
Sweden
Correspondence to: Bengt Glimelius, Department of Oncology, Radiology and Clinical Immunology, Section of
Oncology, University Hospital, SE-751 85 Uppsala, Sweden. Tel: »46186115513 . Fax: »46 186115528 . E-mail:
bengt.glimelius@onkologi.uu.se
Acta Oncol Downloaded from informahealthcare.com by Erciyes Univ. on 12/24/14
Acta Oncologica Vol. 40, No. 2:3, pp. 231 – 252, 2001
A systematic review of chemotherapy trials in several tumour types was performed by The Swedish Council of Technology Assessment
in Health Care (SBU). The procedures for the evaluation of the scienti c literature are described separately (Acta Oncol 2001; 40:
155 –65). This synthesis of the literature on chemotherapy in patients with acute myeloid leukaemia (AML) is based on 129 scienti c
articles; one meta-analysis, 51 randomised trials, 39 prospective and 18 retrospective studies, and 20 other articles. Altogether, 39 557
patients were included in these studies. The conclusions reached can be summarized into the following points:
Standard induction therapy for patients with AML, consisting of daunorubicin and ara-C in conventional doses, results in a complete
remission (CR) rate of 50 – 60% in an unselected population and a long-term survival of about 10 – 20%. The total doses of both ara-C
and daunorubicin are of importance for remission duration and in some studies also for survival.
For personal use only.
High-dose ara-C in the induction therapy prolongs remission duration in randomised trials, but has not been proven to affect
long-term survival. It also increases toxicity and is not generally recommended.
Idarubicin, another anthracyclin, has been compared with daunorubicin in conjunction with ara-C, resulting in a higher CR rate,
especially in younger patients. In a meta-analysis of the ve-randomised trials performed, a slight survival advantage was also seen
with idarubicin. Yet, there is inconclusive evidence to conclude that idarubicin is superior to daunorubicin, and further trials are
needed.
Mitoxantrone improves the outcome of induction therapy in comparison with daunorubicin in some randomised studies, but
conclusive evidence is still lacking.
The addition of etoposide to daunorubicin or mitoxantrone and ara-C has improved CR rates, but has not convincingly improved
survival and secondary leukaemias may be induced.
New induction treatment strategies are de ned by identi cation of prognostic subgroups. A risk strati cation of AML patients as to
chromosomal aberrations might be of importance for the choice of therapy. Moreover, the speed and the morphological response to
the rst induction course are predictive for relapse. However, no prospective randomised studies are as yet published regarding
risk-adapted induction therapy.
Post-remission dose-intensive chemotherapy prolongs the duration of remission, seemingly most in patients B 60 years. However, the
data in support of these conclusions are sparse. A convincing effect on survival has not been shown.
Limited data indicate that post-remission maintenance therapy with long-term attenuated chemotherapy prolongs time to recurrence,
without evidence for prolongation of survival.
Allogeneic bone marrow transplantation is an established practice for consolidation in rst remission for young patients with an
HLA-matched sibling. It is however not known which patients will really bene t from transplantation as no truly randomised
comparison of allogeneic vs autologous transplantation or conventionally-dosed chemotherapy has been performed. Patients with and
without an HLA-identical sibling have been compared on the basis of intention-to-treat principles (‘genetic randomisation’). The
disease-free survival seems to be prolonged in the donor group, due to a lower relapse rate with allogeneic transplantation. A higher
procedure-related mortality makes the effects on total survival uncertain. Randomised trials with autologous transplantation vs
conventional consolidation show a lower relapse rate and a trend for an improved disease-free survival. In one study, in which an
autograft was added to four courses of intensive therapy, there was also a late survival advantage. Thus, the role for intensi ed
post-remission treatment in rst complete remission with high-dose chemotherapy followed by allogeneic or autologous marrow or
1
Other members of the SBU-group were: Jonas Bergh, Radiumhemmet, Stockholm, Sweden; Lars Brandt, Dept of Oncology, University
Hospital, Lund, Sweden; Bengt Brorsson, SBU, Stockholm, Sweden; Barbro Gunnars, Dept of Oncology, University Hospital, Lund,
Sweden; Larsolof Hafström, Dept of Surgery, University Hospital, UmeaÊ , Sweden; Ulf Haglund, Dept of Surgery, University Hospital,
Uppsala, Sweden; Thomas Högberg, Dept of Gynaecological Oncology, University Hospital, Linköping, Sweden; Karl G. Janunger, Dept
of Surgery, University Hospital, UmeaÊ , Sweden; Per-Ebbe Jönsson, Dept of Surgery, Helsingborgs lasarett, Helsingborg, Sweden; Göran
Karlsson, Handelshögskolan, Stockholm, Sweden; Gunilla Lamnevik, SBU, Stockholm, Sweden; Sten Nilsson, Radiumhemmet, Stock-
holm, Sweden; Johan Permert, Dept of Surgery, University Hospital, Huddinge, Sweden; Peter Ragnhammar, Radiumhemmet,
Stockholm, Sweden; Sverre Sörenson, Dept of Thoracic Medicine, Haukeland University Hospital, Bergen, Norway, Sweden.
© Taylor & Francis 2001. ISSN 0284-186 X Acta Oncologica

232 E. Kimby et al. Acta Oncologica 40 (2001)
stem cell transplantation requires further studies. Moreover, studies with strati cation of therapy according to predictors for prognosis
in the individual patient are needed.
Allogeneic stem cell transplantation after minimal or reduced myeloablative conditioning (‘mini-transplantation’ or non-myeloablative
stem cell transplantation) induces a host-vs-graft tolerance and an immune graft-vs-leukaemia effect. This new concept of cellular
immunotherapy seems to have a low procedure-related mortality, but long-term effects are unknown and evaluation in controlled
clinical studies is required.
Patients with relapsed AML can only infrequently achieve long-term remissions with chemotherapy in conventional doses. Uncon-
trolled data indicate that allogeneic transplantation can be a curative treatment for these patients as well as for those refractory to
initial conventional chemotherapy. No studies have compared the effect of allogeneic transplantation in rst compared with second
remission.
Treatment of elderly patients is controversial. In selected elderly patients with good performance status and absence of concomitant
diseases, a combination of ara-C with an anthracycline, both in conventional doses, is the treatment of choice to prolong survival. Oral
cytotoxic drugs, e.g. hydroxyurea and 6-thioguanine, or low dose ara-C subcutaneously are other treatment options that will lead to
remission in a few patients, seemingly with a good quality of life and with few days spent in hospital.
The use of haematopoietic growth factors in patients with AML is doubtful. Most controlled studies show a shortened time for
neutropenia and less infectious complications, but no better effect with reference to remission or survival.
Acute myeloid leukaemia (AML) is an uncommon malig- these, 291 had received modern standard induction ther-
nancy with about 300 new adult patients every year in apy. The median follow-up was 14.7 years. The remission
Sweden. The incidence of AML increases with age and the rate and survival were signi cantly better with standard
median age of adult AML is 60 – 65 years. There are induction therapy for patients with normal cytogenetics or
several accepted risk factors for development of AML as t(15;17). Multivariate analyses showed that karyotype was
chemicals and irradiation but also treatment with cyto- an independent predictor of survival for all patients, i.e.
static drugs, especially alkylating agents and etoposide. also those not receiving standard induction therapy. In
AML is divided into eight main groups according to the another study performed by the Medical Research Council
FAB-classi cation. One of these, promyelocytic leukaemia, (MRC) AML 10 trial (3), including children and adults up
is mostly regarded as a separate entity with speci c thera- to 55 years of age, diagnostic cytogenetics were also an
peutic options and is thus not included in this review. independent prognostic factor. On the basis of response to
The treatment of AML has improved over the last two induction treatment, relapse risk, and overall survival,
decades with more ef cient induction and consolidation three prognostic groups were de ned by cytogenetic ab-
chemotherapy and better supportive measures. Without normalities: AML associated with t(8;21), t(15;17) or
speci c therapy the disease is rapidly fatal. Infections and inv(16) predicted a relatively favourable outcome. In pa-
hemorrhages are major causes of death. Combination tients lacking these favourable changes, the presence of a
chemotherapy results in a complete remission (CR) in the complex karyotype, ¼ 5, del(5q), ¼ 7, or abnormalities of
majority of younger patients. However, despite the use of 3q carried a relatively poor prognosis. The remaining
intensive consolidation therapy, more than half of the group of patients including those with 11q23 abnormali-
patients develop recurrent disease. Treatment of elderly ties, »8, »21, »22, del(9q), del(7q) or other miscella-
patients with AML is controversial, and their survival has neous structural or numerical defects not encompassed by
not improved as much as that of younger patients. the favourable or adverse risk groups had an intermediate
There are several problems how to interpret the results prognosis. The presence of additional cytogenetic abnor-
from trials in AML. The results from many studies repre- malities did not modify the outcome of patients with
sent the outcome of treatment in highly selected popula- favourable cytogenetics. These cytogenetic ndings were
tions of patients. This is especially true for studies also key determinants of outcome from autologous or
regarding transplantation. Moreover, many factors unre- allogeneic bone marrow transplantation (BMT) in rst CR
(see below).
lated to therapy determine the outcome for a patient with
AML. The most in uencial factors on remission induction
INDUCTION THERAPY, CONSOLIDATION,
are age, performance score, tumour burden, with white
MAINTENANCE AND SECOND LINE THERAPY
blood cell counts, serum albumin and lactic dehydrogenase
(LDH). More recently, cytogenetic aberrations have been Role of indi×idual drugs and dose intensity in induction
shown to be of great importance for the outcome of therapy
patients with AML (1–3). In an International Workshop The aim of the initial cytostatic treatment is to achieve a
on Chromosomes in Leukemia, cytogenetic and clinical CR. Without CR the survival is short (usually less than
data from 716 prospectively collected patients with AML one year). In an unselected patient material the CR fre-
were analysed (1). The patients were diagnosed between quency is about 50 – 60% with a standard chemotherapy
1980 and 1982 and 628 patients were de novo AML. Of combination.
Acta Oncologica 40 (2001) Chemotherapy effects in AML 233
Ara-C has been an essential drug in combination A 2 : 1 randomisation was used for induction therapy; 493
chemotherapy since 1968, when Ellison et al. published the patients were allocated to standard-dose and 230 patients
rst results with long-term survivors in AML (4). A steep to high-dose ara-C. The remission rate was slightly, al-
dose-response curve for ara-C has been shown in experi- though statistically insigni cantly, lower with HIDAC as
mental tumour systems, and patients with myeloblasts induction therapy; 55% for patients B 50 years and 45%
retaining higher levels of ara-C-5Æ-triphosphate have higher for patients 50 – 64 years old compared with conventional
CR rates (5). Attempts to improve the clinical effect of doses, 58% and 53%, respectively. Relapse-free survival at
ara-C have included increasing doses. Doubling the dose four years was, however, somewhat better following HI-
from the standard 100 mg:m2 or prolonging the adminis- DAC in the younger age group, 33% vs 21% with conven-
tration time to ten days (from the standard seven days), tional dose (p ¾ 0.049). More toxicity was noted with
both in combination with daunorubicin, has not been HIDAC and survival was not signi cantly prolonged. The
successful (6, 7). Considerably higher doses of ara-C (1 –3 estimated survival at four years for patients below 50 years
g:m2 ) have yielded good results in a number of combina- of age was 32% and 22% in the high and low dose,
tion trials and also as post-induction therapy (see below). respectively (ns). The corresponding gures for older pa-
Most studies have compared induction regimens with tients were 13% and 11% (ns).
modi ed doses of both ara-C and an anthracyclin, which is In summary, both studies found a prolonged remission
why interpretation of the results is dif cult. However, two duration with HIDAC as induction therapy for younger
randomised trials compared high-dose ara-C (HIDAC) AML patients. However, toxicity was increased and the
directly with standard induction regimens with the same value of HIDAC in the induction therapy is still unclear.
postinduction therapy in the two arms (8, 9). This conclusion is probably valid for most cytogenetic
In the rst study (8), 301 patients aged 15 –60 years with subgroups although patients with t(8;21) are suggested to
newly diagnosed AML, were randomised to receive either have a bene t of HIDAC especially when used for consol-
HIDAC (3 g:m2 for 3 hour infusion 12 hourly on days idation (see below) (2).
one, three, ve and seven for eight doses, daunorubicin 50 Daunorubicin is the most widely used anthracycline in
mg:m2 days one to three and etoposide 75 mg:m2 days one AML treatment and induced a CR rate of 30 –50% as a
to seven (HIDAC-3 –7), or standard dose cytarabin 100 single agent (10, 11), and 55 – 75% when used in combina-
mg:m2 continuous infusion for seven days with daunoru- tion with ara-C (12, 13). Increasing the dose of daunoru-
bicin and etoposide at the same doses as above (7– 3–7). If bicin from the standard dose of 45 mg:m2 has not been
a CR was not attained after one course, a second and third studied in prospective comparisons. However, in sequen-
induction course was given. All patients received the same tial studies from the South West Oncology Group
post-induction consolidation therapy for two courses (5– (SWOG), a better response rate was reported with a dose
2 –5). CR was achieved in 71% with HIDAC-3 –7 and in of 70 mg:m2 (14) compared with 45 mg:m2 (9). Moreover,
74% with 7 –3– 7 (ns). For patients attaining CR, the reducing the dose from 45 mg:m2 to 30 mg:m2 resulted in
estimated median remission duration was 45 months with a lower response rate, especially in adult patients B 60
HIDAC-3 –7 and 12 months with 7 –3– 7 (p¾ 0.0005). The years of age (15).
estimated percentage of patients relapse-free ve years Idarubicin, an analogue of daunorubicin demonstrated
after achieving a CR was 49% on HIDAC-3 –7 and 24% good CR rates in patients with refractory AML in phase II
on 7– 3– 7 (p ¾ 0.007). The bene t of the longer remission studies (16– 18) and has also been used as primary therapy.
duration for patients on HIDAC-3 –7 was offset by the Five randomised studies compared the ef cacy of
shorter survival for those who did not achieve CR and in daunorubicin with that of idarubicin in conjunction with
particular by the higher death rate in induction. Totally, ara-C in patients with newly diagnosed AML (19 –23). All
the estimated survival after ve years was 31% for patients trials demonstrated a higher CR rate in patients receiving
with HIDAC-3– 7 and 25% with 7– 3–7 (ns). The HIDAC idarubicin. In four of the studies, these differences were
regimen was signi cantly more toxic with more leukope- statistically signi cant (19, 21 –23). The remission duration
nia, thrombocytopenia, nausea, vomiting and eye prob- and overall survival was prolonged with the new anthracy-
lems. The incidence of central nervous system toxicity, cline in two of the studies (19, 23). A long-term follow-up
including cerebellar, did not differ between the two regi- of survival data from three of the randomised trials (19,
mens. The other study compared HIDAC (2 g:m2 every 12 22, 23) was published in 1997 (24). In one of the studies
h for 12 doses) with conventional doses ara-C (200 mg:m2 (19), in which only patients below the age of 60 were
as infusion for seven days) both for remission induction included, both the originally published survival data and
and for consolidation in patients below the age of 65 years long-term survival favoured the idarubicin group (p¾
(9). Ara-C was combined with daunorubicin at 45 mg:m2 : 0.025 and p ¾ 0.015, respectively), while the early differ-
d for three days in both arms. The trial used a crossover ence in survival in the second study (23) did not remain
design to allow patients on standard-dose induction to signi cant in the updated analysis. In the third study (22),
receive either a high-dose or standard-dose consolidation. no survival advantage was seen in any of the analyses. A
meta-analysis by the Oxford MRC clinical unit (25) used CR was 240 days in patients treated with mitoxantrone
individual patient data from the ve randomised trials and 198 days in those treated with daunorubicin (ns) and
comparing idarubicin and daunorubicin in combination the corresponding gures for median survival were 328
with cytosine arabinoside as induction therapy in 1 052 days and 247 days, respectively (ns). The toxicity pro les
adult patients with newly diagnosed AML. Early induction of the two regimens were comparable.
failures were similar with the two treatments, but after day Mitoxantrone was compared with daunorubicin also in
40, fewer induction failures were seen with idarubicin (17% an EORTC phase III trial (AML 9) in elderly patients (30)
vs 29%; p B 0.0001). CR rates were therefore higher with with the observation of less chemoresistance in patients
idarubicin (62% vs 53%; p ¾ 0.002). Among remitters, treated with mitoxantrone (see below; treatment of elderly
fewer patients allocated to idarubicin relapsed, but slightly patients). The observed trend of better effects with mitox-
more died in remission. Overall survival in these ve trials antrone has led to two additional, still ongoing, ran-
was marginally signi cantly better with idarubicin than domised trials. In the MRC Acute Leukaemia Trial 12, the
with daunorubicin (13% vs 9% alive at ve years; p ¾ aim is to compare a standard regimen, ara-C, daunoru-
0.03). There was a trend for the bene t of idarubicin to bicin and etoposide (ADE) with the mitoxantrone contain-
decrease with increasing age (p ¾ 0.006 for remission rate). ing MAE regimen for patients B 60 years. In both
Totally 569 of the patients were \ 60 years old. The regimens ara-C is given during ten days. For older patients
conclusion of the metaanalysis was that idarubicin was a mitoxantrone-based MAC regimen, which includes ara-C
better with respect to CR rates and survival than daunoru- during ve days, is compared with ADE (MRC AML 10).
bicin for induction in younger patients with de novo Doxorubicin has been used for treatment of AML pa-
AML. The true difference between daunorubicin and tients in a few studies but with a suggested higher toxicity,
idarubicin if given at equitoxic doses has been debated. In mainly mucositis and other gastrointestinal toxicity (15).
none of the above ve randomised trials was the haemato- In a randomised study including 100 patients, mentioned
logical toxicity of induction therapy signi cantly greater in the meta-analysis described above, the effect of doxoru-
with idarubicin. However, in two of the studies (22, 23), bicin was compared with that of idarubicin without differ-
haematologic toxicity was greater with idarubicin during ences (25). DNA-bound doxorubicin in combination with
the consolidation phase. Thus, new prospective trials of ara-C, thioguanine, vincristine, and prednisolone
idarubicin in the doses used above vs daunorubicin at (POCAL) was used in a Swedish randomised study (in-
higher doses (\ 50 mg:m2 :day for three days) might be of cluding totally 120 patients) (31). The study showed better
interest. effect as to remission duration and survival of the doxoru-
The synthetic antracenedione-derivative mitoxantrone bicin-DNA containing regimen compared with conven-
has also been effective in AML based on phase II studies tional doxorubicin (p B 0.025 and B 0.01, respectively).
(26–28). The results from a multi-centre randomised trial This regimen was not, however, con rmed to be effective
suggested that mitoxantrone had an advantage compared in a later study from the same group involving 86 patients
with daunorubicin (29). This phase III trial in previously (32). This study was closed due to a low remission rate in
untreated adults compared mitoxantrone plus ara-C with the POCAL-arm. The difference in CR rate with POCAL-
the Cancer and Leukaemia Group B (CALGB) ‘7 »3’ DNA between the two studies remains unexplained, but
daunorubicin-based regimen. Two hundred evaluable pa- the source and preparation of the DNA were different,
tients (98 treated with the mitoxantrone-based regimen which might be of importance for its ef cacy. Both dox-
and 102 with the daunorubicin-based regimen) were in- orubicin and daunorubicin are today available in a liposo-
cluded in an analysis of ef cacy. The median age was 60 mal form, which might reduce the cardiotoxicity.
years. The induction regimen comprised ara-C 100 mg:m2 However, no randomised studies have been published as to
by infusion daily for seven days and mitoxantrone 12 antileukaemic effects of the bound anthracyclines.
mg:m2 or daunorubicin 45 mg:m2 daily for days one to The rationale for the use of etoposide in acute
three. If needed, a second induction course was adminis- leukaemia came from in vitro studies, which showed syn-
tered: ara-C for ve days and mitoxantrone or daunoru- ergy between etoposide and ara-C (33). When etoposide
bicin for two days. Postremission therapy consisted of two was added to remission induction therapy in newly diag-
consolidation courses, identical to the second induction nosed patients, results were encouraging (34). However,
course. Sixty-three% of patients treated with mitoxantrone the use of etoposide has been associated with an increased
achieved CR compared with 53% treated with daunoru- risk of secondary leukaemias.
bicin (p¾ 0.15). The median time to CR was 35 days in The Leukaemia Study Group of Australia compared
patients treated with mitoxantrone and 43 days for those daunorubicin (50 mg:m2 daily i.v. bolus for three days)
treated with daunorubicin. This difference re ects the fact »ara-C (100 mg:m2 daily continuous infusion during
that CR occurred more frequently after a single induction seven days (‘7–3’) with the same regimen with the addition
course of the mitoxantrone-based regimen (89%) compared of etoposide (75 mg:m2 daily for seven days) (‘7 –3– 7’) in
with the standard regimen (68%). The median duration of a randomised trial including 264 eligible patients (34). CR
occurred in 56% of 7– 3 and 59% of 7–3-seven patients with one of three different chemotherapy regimens;
with a median remission duration of 12 months for 7 –3 thioguanine »Ara-C » Daunorubicin (TAD), Daunoru-
and 18 months for 7– 3– 7 (p¾ 0.01). Survival was similar bicin »Ara-C (DA) or Ara-C »vincristine »methotrex-
between the two groups. Subset analyses performed to ate »leucovorin (AVML). The latter ‘‘synchronizing-
identify patients with the most bene t showed that recruiting’’ drug schedule was used with the intention to
etoposide signi cantly prolonged remission duration in manipulate the cell cycle instead of giving a maximal
younger patients (less than 55 years) with a median of 12 cytotoxic effect. Patients attaining CR were consolidated
months for 7 –3 and 27 months for 7– 3– 7 (p¾ 0.01). Also with TAD or AVML and maintenance treatment was
survival was prolonged by 7 –3– 7 in patients aged less given with Bacillus Calmette-Guerin (BCG) vaccination,
than 55 years, with a median of nine months for 7 –3 as or BCNU plus Ara-C (B:A), or no further therapy (NFT).
compared with 17 months for 7– 3– 7 (p¾ 0.03). In older Of 209 evaluable TAD patients, 105 (50%) achieved CR;
patients 7– 3–7 was more toxic, with signi cantly more and of 187 DA patients, 97 (52%) achieved CR while
severe stomatitis (p ¾ 0.02) and no additional clinical AVML yielded only 15 CR among 59 patients (25%). The
bene t. time to remission was signi cantly shorter with DA com-
In a randomised study from the Swedish Leukaemia pared with TAD. Thus, no early advantage with the
Group (32), also described above, mitoxantrone, addition of thioguanine was noted and the long-term
etoposide, and ara-C (MEA) was compared with doxoru- effects were not evaluated as patients were randomized to
bicin-DNA, ara-C, thioguanine, vincristine, and pred- three options for maintenance therapy.
nisolone (POCAL-DNA). The study was closed after an
interim analysis showing that the MEA regimen had a very Dose intensity in combinations in induction therapy
high antileukemic activity; CR in 35:42 (83%) vs 20:44 The dose and dose intensity of an anthracycline in combi-
(45%) in the POCAL-DNA treated patients (pB 0.001) nation with ara-C have in uenced the duration of remis-
and a median survival of 28 vs 13 months (pB 0.03). sion in AML according to several studies. A large data
The relative ef cacy and toxicity of etoposide was com- base from the Australian Leukaemia Study Group
pared with thioguanine in a randomised comparison of the (ALSG) showed that the induction dose of both ara-C and
ADE (ara-C, daunorubcin, etoposide) vs DAT (daunoru- daunorubicin was of signi cant importance for remission
bicin, ara-C, thioguanine) regimen in the MRC Acute duration (39). Also, the conclusion from the MRC AML 9
Leukaemia Trial 10 (35). Totally, 1 857 eligible patients, trial was that a more intensive induction regimen DAT
most of them less than 56 years old were randomised. The 3 »10 is more effective than DAT 1 »5 for all age groups
two treatment groups were well matched for presentation (40). In this trial, 972 patients were randomised between
features. CR rate was 81% with DAT and 83% with ADE DAT 1 »5 (daunorubicin for one day, cytarabine and
(ns). Patients receiving DAT took slightly but statistically 6-thioguanine for ve days) and DAT 3 »10 (same dose of
signi cantly longer to recover from neutropenia and drugs for three and ten days, respectively). Resistance to
thrombocytopenia. However, the median numbers of days induction therapy was less common with the DAT 3 »10
in hospital were similar in the two groups. ADE patients regimen (13% vs 23%; p¾ 0.0001) and the CR rate was
experienced slightly more severe non-hematologic toxicity. slightly higher compared with DAT 1 »5 (66% vs 61%;
The death rate in CR during consolidation chemotherapy p ¾ 0.15). Moreover, CR was achieved more rapidly with
was 9% with ADE compared with 6% with DAT (p ¾ DAT 3 »10 (median 34 vs 46 days; pB 0.0001). The
0.06). No statistically signi cant differences between the ve-year relapse-free survival and overall survival were
two treatment groups were seen regarding disease-free also better with DAT 3 »10 (28% vs 23%; p ¾ 0.05, and
survival, relapse rate or survival at seven years (survival 23% vs 18%; p B 0.05, respectively). However, DAT 3 »10
40% for both DAT and ADE). Also subgroup analyses was more toxic with a 5% increase in the risk of induction
failed to show any bene t for etoposide in patients with death.
monocytic or myelomonocytic disease or in any other In a three-armed randomised trial from 15 institutions
diagnostic subgroup. In conclusion, DAT and ADE both in France a ‘3 »7’ chemotherapy (group 1: daunorubicin
achieve high remission rates and good long-term survival 80 mg:m2 d 1 –3 » ara-C 200 mg:m2 d 1 –7) was compared
and are equally effective chemotherapy regimens for the with a double induction and a sequential induction
treatment of AML patients up to 55 years (35). chemotherapy (41). Totally 307 adult patients less than 50
In an early AML study (36) thioguanine was added to years with de novo AML were included. The double
ara-C and a positive effect on remission rate was sug- induction (group two) consisted of daunorubicin 80 mg:m2
gested. However, in two phase II studies, the addition of d 1– 3 and ara-C 200 mg:m2 d1– 7 »mitoxantrone 12
6-thioguanine did not appear to improve the overall re- mg:m2 d 20 –21 »ara-C 500 mg:m2 :12h d20–22. The se-
sults, that were achieved with a combination of anthracy- quential induction (group 3) consisted of daunorubicin 80
cline and ara-C (37, 38). In the rst of these studies (37), mg:m2 d 1–3 »ara-C 200 mg:m2 d 1– 3 »mitoxantrone 12
patients were randomly allocated to remission induction mg:m2 d8–9 »ara-C 500 mg:m2 :12h d 8 –10. One course
of salvage therapy was given to patients who were not in randomly assigned to receive four courses of ara-C at one
remission after one induction course. All patients received of three doses: a) 100 mg:m2 ve days by continuous
the same salvage and the same two consolidation courses. infusion b) 400 mg:m2 ve days by continuous infusion or
The CR rate was 83% with no differences between the c) 3 g:m2 in a 3-hour infusion every 12 hours (twice daily)
groups. However, for 18% of the patients in group 1 and on days one, three and ve. All patients received four
15% in group 3, one salvage course was needed to reach a courses of monthly maintenance treatment. The probabil-
CR compared with only 4% in group 2 (p¾ 0.007). Three- ity of remaining in continuous CR after four years for
year disease-free survival was 50% and 52% in groups 2 patients 60 years of age or younger was 24% in the 100
and 3, respectively, and 40% in group 1 (ns). Toxic deaths mg-group, 29% in the 400 mg-group, and 44% in the 3
and infectious complications were not different between g-group (p ¾ 0.002). In contrast, for patients older than 60,
the groups. Thus, sequential induction and double induc- the same probability was 16% or less in each of the three
tion chemotherapy were safe but it is still unclear whether post-remission groups. The data support the concept of a
there is a therapeutic advantage compared with a conven- dose-response effect of cytarabin in younger patients.
tional ‘3 »7’ chemotherapy even in young AML patients. The randomised SWOG-study (9) compared HIDAC (2
g:m2 every 12 h for 12 doses) with conventional doses
Post -remission consolidation therapy ara-C (200 mg:m2 :d as infusion for seven days) both for
Different strategies have been used to prevent relapse in remission induction and for consolidation in combination
AML patients in rst remission. Post-remission intensi ca- with daunorubicin at 45 mg:m2 :d for three days. Complete
tion of therapy has been evaluated in several studies. In a responders to standard-dose ara-C were randomised to
report published by Jehn (42), 122 adults (15 –65 years) receive either two additional courses with the same dose
were recruited to two sequential prospective phase II stud- (n¾ 126) or one high-dose ara-C (n¾ 119). No advantage
ies involving a 3 –7 –type induction regimen. Progressive as to disease-free or total survival was seen with high dose
increments in the intensity of post-remission therapy ara-C (p¾ 0.77 and 0.46, respectively). All patients in CR
seemed to yield an improvement of remission duration and with high-dose ara-C as induction were scheduled for
survival. high-dose also as consolidation. In a non-randomised com-
High dose ara-C as consolidation has led to a prolonged parison, the patients receiving both high-dose ara-C induc-
leukaemia-free survival in some young patients with AML tion and consolidation seemed to have the best survival.
in phase III studies (43, 44). The Eastern Cooperative However, there seemed to be a selection bias as only 77%
Oncology Group (ECOG) randomised 193 patients with- of the patients in CR in the high dose arm received
out an allogeneic donor in CR after one or two induction high-dose consolidation, most probably due to toxicity.
courses with three days of daunorubicin 60 mg:m2 , ve The Swiss Group for Clinical Cancer Research (SAKK)
days of continuous infusion ara-C 200 mg:m2 and ve performed a trial to evaluate the effectiveness of one single
days of thioguanine to one of three treatment arms (43). post-remission course of HIDAC (46). Adults with de
The observation arm (no consolidation) was closed after novo AML (15 –65 years) in remission after two induction
an interim analysis because of inferior remission duration courses were randomly assigned to one consolidation
(45). The consolidation therapy consisted of one course course either with standard ara-C (SDAC: 100 mg:m2 24
(ara-C 3 g:m2 over one hour every 12 h for 12 doses hours infusion over seven days) or with high-dose ara-C
followed by amsacrine 100 mg:m2 days 7– 9), while in the (HIDAC: 3 g:m2 every 12 hours as one-hour-infusion for
maintenance arm therapy was continued for two years six days). In addition, both arms included daunorubicin
(thioguanine 40 mg:m2 orally every 12 hours for 4 days (45 mg:m2 :d, days 1–3). Thereafter, patients were ob-
followed by ara-C 60 mg:m2 subcutaneously on the fth served without maintenance until relapse. After the two
day every week). The calculated four-year event-free sur- induction courses, 208:276 eligible patients achieved remis-
vival and survival for patients in the consolidation arm sion (CR: 61%; PR: 4%), 41 were resistant (15%) and 20
was 27 and 33%, respectively, compared with 16 and 22% died early (7%). Seventy-one patients in remission were not
in patients with maintenance (ns). In a subgroup analysis randomised. One hundred and thirty-seven patients were
in patients B 60 years, the four-years event-free survival randomised in CR:PR (67 SDAC, 70 HIDAC). WHO
was 28% in the consolidation arm compared with 15% in grade 3 –4 toxicities occurred in 14:67 SDAC and in 38:66
patients with maintenance (p¾ 0.047). The consolidation HIDAC patients (p B 0.0001). The median event-free sur-
with high dose ara-C seemed to be of value only in vival was 10.8 (SDAC) vs 12.2 months (HIDAC; p ¾ 0.18)
younger patients, mainly due to a high mortality rate in and the median overall survival was 24.6 (SDAC) vs 32.6
patients \ 60 years. months (HIDAC) (p¾ 0.07). For 112 patients strati ed as
A study from the CALGB (44) included 1 088 adult CR, the estimated four-year disease-free survival was 25%
patients with newly diagnosed AML. Induction therapy with SDAC and 37% with HIDAC (p¾ 0.09). The overall
was given with three days of daunorubicin and seven days survival rates at four years were 38% and 48%, respectively
of cytarabine. Of the 693 patients who had a CR, 596 were (p¾ 0.10). The authors concluded that early consolidation
of adult AML in CR with a single course of HIDAC ages of 15 and 50 were given daunorubicin, 45 mg:m2 , for
seemed to be superior in terms of outcome, although not the rst three days of a ten-day continuous infusion of
statistically signi cant, to one cycle of SDAC. ara-C (initially a dose of 200 mg:m2 , but reduced to 100
The relationship of intensi cation with cytarabine to mg:m2 because of toxicity). Those under the age of 36
outcome by the cytogenetic group was evaluated in 285 years achieving a CR and with a histocompatible donor
newly diagnosed patients with primary AML (2). All pa- were assigned to a transplant arm. The rest of the patients
tients had adequate karyotypes and were enrolled on a were randomised to one of three consolidation arms: a)
prospective CALGB cytogenetic study. The patients were ara-C 200 mg:m2 daily for seven days and daunorubicin 45
randomly assigned to postremission treatment with stan- mg:m2 daily for three days for three courses; b) one course
dard-, intermediate-, or high-dose cytarabine intensi ca- as in arm a, followed by amsacrine 120 mg:m2 daily for
tion. Patients were categorized to one of three cytogenetic ve days followed by a ve-day continuous infusion of
groups: (a) core binding factor type [(CBF); i.e. t(8;21) azacytidine 150 mg:m2 daily; c) thioguanine and ara-C 100
inv(16), t(16;16), and del(16)]; (b) normal; and (c) other mg:m2 every 12 hour and daunorubicin 10 mg:m2 daily for
abnormality karyotype. After a median follow-up time of ve days for the three courses followed by four mainte-
more than seven years the treatment outcome of CBF nance courses of ara-C (100 mg:m2 daily for ve days and
AML patients was superior, with an estimated 50% still in daunorubicin 45 mg:m2 for two days every 13 weeks).
Totally, 398 patients were eligible and 219 achieved a CR.
CR at ve years as compared with 32 and 15% for patients

with normal karyotype AML and other abnormality Only eleven patients were assigned to the transplant arm.
AML, respectively (pB 0.001). Univariate analyses showed The ve-year disease-free survival was 38%, 31%, and 27%
that a higher postremission cytarabine dose (pB 0.001) in consolidation arms A, B, and C, respectively (ns). Thus,
in this study the intensive consolidation therapy with the
was of value for prolonged CR duration. The impact of
same or different drugs as used in induction was not more
cytarabine dose on long-term remission was most marked
effective than lower dose consolidation followed by
(pB 0.001) in the CBF AML group with 78% of those
maintenance therapy.
with a dose of 3 g:m2 still in CR at ve years. The
corresponding gure for those with a dose of 400 mg:m2
Post -remission maintenance therapy

was 57% and for those with a dose of 100 mg:m2 only
Modern post-remission therapy mostly includes consolida-
16%. For patients with a normal karyotype, 40% of those
tion (see above). Also attenuated chemotherapy during
with a dose of 3 g:m2 were still in CR as were 37% of those
many months as ‘‘maintenance’’ can prolong CR duration
with a dose of 400 mg:m2 and 20% of those with a dose of
compared with no further therapy (45). A long-term obser-
100 mg:m2 . In contrast, cytarabine at all doses produced
vation of patients from the German AMLCG 1981 study
only a 21% or less chance of long-term continuous CR for
was recently presented (48). The standard induction and
patients with other cytogenetic abnormalities.
consolidation with TAD and monthly maintenance during
Besides ara-C, other drugs or combinations of drugs
three years showed a probability of disease-free survival of
have been used for consolidation in AML patients. In the
23% and 20% at ve and ten years, respectively, while in
MRC AML 9 trial (40), the 63% of the patients who patients without maintenance the ve-year disease-free sur-
achieved CR were randomised to receive two courses of vival was only 5%. The CALGB studied two schedules for
DAT 2 »7 alternating with two courses of either MAZE duration of maintenance; eight months compared with
(m-AMSA, 5-azacytidine, etoposide) or COAP (cyclophos- three years, without any difference in CR duration (38).
phamide, vincristine, ara-C, prednisone). Post-remission The value of maintenance only compared with intensive
intensi cation of therapy with MAZE resulted in fewer consolidation was also evaluated in the randomised study
relapses (665 vs 745 at ve years; p ¾ 0.003), but the by the German AMLCG group (48). Maintenance
patients required more supportive care. Totally, 4.5% of chemotherapy vs intensive consolidation or conventional
the patients died following MAZE, while no deaths oc- consolidation was evaluated in 321 patients in CR follow-
curred following COAP. Thus, intensive post-remission ing intensive induction. One course of a sequential high
therapy with MAZE achieves a better leukaemic control dose ara-C and mitoxantrone (S-HAM) as intensive con-
but with substantial toxicity. The ve-year survival was solidation appeared equivalent in its antileukemic activity
not statistically signi cantly higher with MAZE (37% vs to prolonged maintenance (monthly courses of standard
31%). dose ara-C-daunorubicin, ara-C-thioguanine, ara-C-cy-
Vogler et al. (47) from the Southeastern Cancer Study clophosphamide and ara-C-thioguanin repeated during
Group published results from a phase-III clinical trial three years). However, since the myelotoxic effect of S-
evaluating if both more intensive induction and consolida- HAM was a problem, this regimen will be supported by
tion therapy increased the remission rate and prolong autologous stem cell infusion in a presently ongoing study.
survival. A minor objective was to determine if the use of The Southwest Oncology Group randomised 150 eligible
non-cross-resistant drugs was more effective for induction. patients to receive late intensi cation alone or late inten-
Patients with previously untreated leukaemia between the si cation plus monthly low-dose maintenance (14). In mul-
tivariate analyses, treatment with low-dose maintenance High-dose Ara-C has been used as a salvage regimen as
was associated with prolonged disease-free survival (p ¾ it may overcome resistance in leukaemic cells (53, 54).
0.028), but not improved overall survival (p¾ 0.27). High dose ara-C (3 g:m2 :d for 6 d) with or without
The value of additional low dose maintenance therapy etoposide (100 mg mg:m2 d 7, 8, 9) was evaluated in a trial
after intensive consolidation was also questioned in the of relapsed or refractory AML (55). Of 67 patients ran-
MCR AML 9 study (40). Patients still in CR after post- domised to ara-C, 31% obtained CR compared with 38%
consolidation therapy were randomised to receive either in the combination group (ns). The median remission
one year of maintenance treatment with eight courses of durations were 12 and 25 months, respectively (ns). The
cytarabine and thioguanine followed by four courses of addition of etoposide to high dose ara-C seemed to have a
COAP, or no further cytotoxic therapy. The low level marginal effect at the expense of an increase in toxicity.
maintenance therapy appeared to delay recurrence but However, for patients under the age of 50 a statistically
conferred no advantage in survival and was found incon- signi cantly longer survival was seen with the combination
venient and costly. (p¾ 0.04).
During recent years, immuno-modulating agents have The German AML CG evaluated the effect of sequential
been used as adjuvant to standard chemotherapy mainte- high-dose ara-C and mitoxantrone (S-HAM) in 186 pa-
nance (49), or as the only therapy (50). The clinical results tients with relapsed or refractory AML (56). Ara-C was
are not mature enough to draw any conclusions from. administered by three hr infusion q 12 hrs on days 1, 2,
and 8, 9 at randomly assigned doses of either 3.0 vs 1.0
g:m2 in patients B 60 years of age or 1.0 vs 0.5 g:m2 in
Treatment of primary resistant AML and second-line older patients. Mitoxantrone was given at a dose of 10
therapy mg:m2 :day on days 3, 4, 10 and 11. Randomisation was
Patients not obtaining a remission on conventional ther- strati ed for primary refractoriness and for the length of
apy or who relapse after a short length of remission have rst remission in relapsed cases ( B six months, 6 –18
a grave prognosis. Thus, it is important to decide whether months, \ 18 months). Seventy-two patients (48%)
intensive treatment with the aim of obtaining a prolonged achieved a CR, 38 cases (25%) were non-responders (NR)
remission-free period or a palliative approach should be and 41 patients (27%) died within the rst six weeks after
used. Patient-related factors are then of great importance. the start of treatment. No signi cant differences were
Karyotype at diagnosis in uences the prognosis also at found in CR rates, being 52% and 44% for the 3.0 vs 1.0
relapse. g:m2 ara-C regimens in patients B 60 years, and 48% and
The probability of achieving a second remission is 45% for the 1.0 vs 0.5 g:m2 ara-C in older patients. In the
highly dependent on the duration of the rst remission total patient material no differences emerged either for the
(51). In this retrospective single-centre study of 168 pa- time to CR (median 46 days) or for remission duration
tients, 66 (39%) patients achieved a second CR. The (median 4.5 months). However, analysis of treatment fail-
probability of a continuous CR in patients with a rst ures demonstrated a signi cantly higher rate of non-re-
remission of more than 12 months was 35% at three years sponders after the lower dose regimens in both age groups,
and 24% at ve years, while none of the patients with a 41% and 32% vs 11% and 14%, respectively in patients
rst remission less than 12 months stayed in remission at receiving ara-C at higher doses (p B 0.01). Thus, the treat-
three years. Median remission duration of those with a ments showed good antileukemic effect, but there was a
long rst remission was 18 months. In this study, duration high incidence of early deaths during treatment-associated
of the second CR was longer than that of the rst one in aplasia, mostly due to infections. In a subsequent trial,
22% of the patients. The reasons for the longer duration of G-CSF (5mg:kg:day sc) was started immediately after the
the second CR that the rst one in a few patients was end of S-HAM therapy and continued until the end of
unclear. One possible explanation is that chemotherapy at neutropenia. In an interim report, the rate of early deaths
relapse often can be initiated at an earlier time point than was reduced, and more CRs and a prolonged time to
at the rst manifestation of AML. However, also the treatment failure were seen in evaluable younger patients
ef cacy of the treatment can be related to response rate [German AML group, personal communication]. In an-
and duration. other study using a very intensive salvage regimen (time
Many different chemotherapy regimens are effective in sequential mitoxantrone, etoposide, and cytarabine), the
the relapse situation. All clinical studies in the English ve-year survival from second CR was 25% in patients
literature from 1976 to mid-1987 that addressed drug with a rst CR of more than six months, while in patients
management of refractory or relapsed ANLL were re- primary refractory or with an early relapse, the survival
viewed by Welborn et al. (52) with the conclusion that was only 12% (57). The corresponding gures for disease-
many regimens may induce a CR, but that none was free survival were 20% and 3%, respectively. Also Kantar-
clearly superior. However, during that time period combi- ijan et al. (58) have shown that patients with a late relapse
nations including etoposide had not been evaluated. (in this study de ned as rst CR duration of more than 18
months) have a chance of a sustained second CR with 25% The literature shows that:
of non-transplanted patients showing a continuous CR at
Standard induction therapy for patients with newly
three years.
diagnosed AML is daunorubicin and ara-C. The total
Amsacrine has been shown to be effective as salvage
doses of both ara-C and daunorubicin are of impor-
therapy either as a single agent (59) or in combination with
tance for remission duration and in some studies also
ara-C (60). In the latter study, amsacrine was combined
for survival. A higher CR rate and a longer remission
with high-dose cytosine arabinoside in 40 relapsed or
duration have been shown for idarubicin compared with
refractory patients with a resulting 70% CR rate. Nine
daunorubicin (both in conjunction with ara-C), espe-
patients died with hypoplastic marrow, ve with less than cially in young patients, and there is a survival advan-
5% blast cells. The median duration of remission for tage of idarubicin according to a meta analysis.
complete responding patients exceeded eight months. Am- High-dose ara-C in the induction therapy prolongs re-
sacrine has also been used in combination with ara-C and mission duration, but has not affected long-term sur-
etoposide for newly diagnosed patients not entering remis- vival and it increases toxicity.
sion after one course of idarubicin-ara-C (61). The am- Other agents, such as mitoxantrone and etoposide, im-
sacrine combination was found to be effective and after prove the outcome of induction therapy in some ran-
two courses, 90 patients (76%) attained a remission. Am- domised studies. The addition of etoposide has
sacrine was also used for consolidation with mitoxantrone, improved CR rates and, in one large study, also sur-
etoposide and ara-C. Totally 81% of the patients went into vival, but conclusive evidence is lacking
remission, of whom 15 patients underwent allogeneic Post-remission dose-intensive chemotherapy prolongs
transplantation and 30 (below the age of 56 years) au- the duration of remission, seemingly most in patients
tologous transplantation in the rst CR. The overall prob- B 60 years. Increased remission duration might also be
ability of survival for all patients was 34% at four years. achieved by long-term attenuated maintenance
Idarubicin has been used and found effective in the chemotherapy. However, the data in support of these
treatment of patients with refractory or recurrent AML conclusions are sparse. A convincing effect on survival
(16, 18, 62). Combination regimens including udarabine, has not been shown.
cytosinarabinoside and G-CSF (FLAG) or the same drugs In patients with relapsed AML, remissions can be at-
with mitoxantrone (FLANG), have also a documented tained in 30 –50% of patients with a duration mostly in
effect on CR rates in patients with refractory leukaemia the range 3– 25 months. However, some patients main-
(63, 64). tain the remission considerably longer, especially those
Low dose therapy can have palliative effects in some with a long duration of rst remission. In the latter
patients with a relapse from AML. In one report, however, patients, it is likely that cure can be obtained in some
only including ten patients, a CR was obtained in six patients. The attaining of a remission might, thus, lead
patients treated with ara-C subcutaneously 10 mg:m2 every to prolonged survival time, but no randomised studies
12 hours for 21 days (65). have been performed and thus it is impossible to quan-
titate the survival gain. Some relapsed patients may get
Resistance re×ersal by cyclosporin analogues palliative effects of low dose therapy.
The over-expression of the multidrug resistence gene is a
major mechanism of chemotherapy resistence in AML. To Induction therapy, consolidation, maintenance and 2nd line
reverse overexpression of the gene product pgp-170 cy- therapy:
closporin analogues have been used (66). No phase III
trials as to the clinical effect in relapsed AML patients Scienti c evidence*
have yet been presented.
1 ¾ High 2 ¾Moderate 3 ¾Low Total
In summary, there are several clinically valuable options Number of studies:number of patients
of therapy for patients refractory to induction chemother-
M 1:1 897 – – 1:1 897
apy or with relapse. Therapy leading to a second remission C 8:5 567 17 :3 746 6:1 203 31:10 516
probably leads to a better quality of life with prolonged P 4:1 556 13 :763 5:156 22:2 475
survival time, but no randomised studies have been per- R 1:1 612 2: 451 1:58 4:2 121
formed. Therefore it is not possible to properly quantitate L 6 – – 6
O 5 – – 5
the survival gain with any chemotherapy regimens. As will
Total 25:10 632 32 :4 960 12 :1 417 69:17 009
be discussed below, it is likely that the chance for cure is
higher if an allogeneic transplantation is performed. For *The weight of scienti c evidence for each publication was graded
as described (Acta Oncol 2001; 40: 155 – 65). M ¾meta-analysis,
the majority of older patients, and those without a suitable C¾ controlled clinical trial, P¾prospective trial, R ¾ retrospec-
donor, both new agents and new strategies are required for tive study, L ¾literature review and O ¾other studies. The clas-
long-term survival. si cation of each publication is given in the reference list.
ALLOGENEIC AND AUTOLOGOUS activity can be induced also in autologous stem cell trans-
HAEMATOPOIETIC STEM CELL plantations, but evidence is lacking regarding clinical ef-
TRANSPLANTATION fects. Autologous stem cell transplantation may result in a
reduction of relapses compared with conventional
Combination chemotherapy for AML results in CR in the
chemotherapy (78 –80). However, potential residual
majority of young patients, but despite the use of intensive
leukemic clonogenic cells in the harvest are reintroduced to
chemotherapy consolidation, more than half of the pa-
the patient with the stem cells. A strategy to minimize this
tients develop recurrent leukaemia. High-dose therapy fol-
risk of relapse is to ‘purge’ the stem cell harvest. The
lowed by transplantation with stem cells, both autologous
results from a multicentre EBMT group study of 263
and allogeneic, is another method of intensifying consoli-
patients autotransplanted in early CR1 after purging with
dation to increase cure rates. Analyses of large retrospec-
4-hydroxycyclophosphamide suggested a decreased relapse
tive patient cohorts have indicated that increased cure
rate (81). However, no randomised studies of purging have
rates are seen if patients are transplanted in rst remission
been performed in AML.
(67, 68). Such analyses have also shown that some patients
The best time point for autologous transplantation has
can be cured by high-dose therapy after relapse (67– 69).
been debated. In a retrospective study the relapse rate was
Myeloablative therapy supported by allogeneic bone
higher when autologous bone marrow stem cells were
marrow transplantation (BMT) has been established prac-
harvested after one consolidation treatment compared with

tice for more than 15 years. It is however not known,
after two or more consolidations (82). This was explained
which patients will really bene t from transplantation as
by a possible reinfusion of more residual leukemic cells
no truly randomised comparisons of allogeneic vs au-
with the autograft and:or more residual tumour after only
tologous transplantation or conventionally dosed
one chemotherapy consolidation.
chemotherapy have been performed (see below).
Haematopoetic stem cells from peripheral blood, col-
The risk of relapse is markedly reduced with allogeneic
lected by apheresis after mobilisation with haematopoetic
transplantation compared with chemotherapy or au-
growth factors and:or cytotoxic drugs, can replace bone
tologous transplantation because of alloreactivity of donor marrow. The use of peripheral stem cells reduces the time
cells against residual leukaemia, the so-called graft-vs- to haematological recovery and decreases the incidence of
leukaemia (GVL) effect (70, 71). A reactivity also against infections in autologous transplantation. In a retrospective
normal recipient cells may lead to graft-vs-host disease EBMT register study including 1 422 AML patients allo-
(GVHD) and increased risk of transplant-related mortality transplanted in rst remission, the clinical outcome did not
(TRM). However, a mild (Grade 1) acute graft-vs-host relate to the source of transplanted stem cells (83). Ran-
disease seems to give the highest leukaemia-free survival domised trials comparing peripheral blood and bone mar-
(72). An optimal myeloablative conditioning chemo-radio- row as source of allogeneic stem cells are currently
therapy, a good pharmacological prophylaxis to reduce ongoing.
GVHD and effective therapy for infections are mandatory
for good results of allogeneic transplantation. The success Comparisons of allogeneic ×s autologous transplantation ×s
of allogeneic transplantation procedures also depends on con×entional chemotherapy
matching of genes within the major histocompatibility No truly randomised comparison of allogeneic transplan-
complex. New techniques for HLA typing at the molecular tation with autologous transplantation or chemotherapy
level are of importance for excluding mismatches with has been performed. However, patients with and without
increased risk of GVHD and rejection and reduced sur- an HLA-identical sibling have been compared (‘genetic
vival after transplantation. randomisation’). There are methodological problems to
Most experience with allogeneic transplantation comes compare the different therapeutic options as only patients
from HLA-matched sibling transplants, but transplanta- remaining in CR for a suf cient period of time can be
tion with stem cells from HLA-matched volunteer unre- included in transplantation arms. Also less t patients are
lated donors is feasible with acceptable overall results (73), excluded from transplantation. As a result of this with-
in young patients not very different from what is reported drawal of patients with poor prognosis the outcome of
for matched siblings (74). However, retrospective EBMT transplantation may appear improved. Moreover, the
data have not shown any advantage of matched unrelated event-free survival for the patients receiving myeloablative
donor transplantation compared with autologous trans- therapy is often calculated from the time of the myeloabla-
plantation, not even in patients with relapse (75). tive therapy rather than from the time of CR, which makes
Autologous bone marrow transplantation has been used comparisons with conventional chemotherapy schedules
with fewer acute problems compared with allogeneic trans- dif cult.
plantation and with a lower TRM, but leads to an in- The problems of patient selection are illustrated in a
creased relapse rate due to the absence of a GVL effect. By prospective registration study conducted by the EBMT
the use of interleukin-2 or cyclosporin (76, 77) a GVL-like group (84). Newly diagnosed patients with AML were
registered at the time of HLA-typing, a timepoint when patients were analyzed according to age (adults or chil-
often a speci c therapy is preselected. This time point is dren) and status at transplants (CR1 or CR2). One thou-
comparable for most patients and can be used for all sand one hundred and fourteen adult patients were
patients regardless of the actual treatment given and thus transplanted in CR1; 516 received an allograft and 598 an
reduces selection biases. The intention to treat in case of autograft. Following allogeneic BMT, the TRM was statis-
presence or absence of an HLA-identical donor was tically signi cantly higher (27 vs 13%; p B 0.001), the
recorded at the time of HLA-typing. Among 79 HLA- relapse incidence (RI) lower (25 vs 52%; pB 0.0001) and
typed patients at diagnosis, 27 had an identical donor the leukaemia-free survival (LFS) better (55 vs 42%; p¾
identi ed (34%). The estimated survival at three years 0.006) compared with ABMT. An updated analysis from
from HLA-typing was 44% and 21% among patients with the EBMT register on totally 14 487 adult patients with
or without a donor, respectively (p ¾ 0.02). Out of 26 acute leukaemia has been presented during the 1997
patients intended for allogeneic transplantation, 22 (85%) EBMT meeting (83). The risk of relapse was lower in the
were transplanted, compared with only 15 (32%) out of 47 2 753 adult patients treated with an allogeneic transplanta-
patients planned for an autotransplantation. Sixty-eight tion in rst CR (25%) compared with 2 431 autografted
patients were HLA-typed after attaining rst remission; patients (51%) and leukaemia-free survival at ve years
the estimated three years-survival for the 40 patients with was longer (54%) compared with those with autologous
a HLA-identical donor was 42% compared with 35% for transplantation (41%).
28 patients without a donor (n.s.). Among the patients These EBMT data on leukaemia-free survival in patients
typed in rst CR, 38 (95%) out of 40 patients with a donor with allogeneic transplantation are comparable with those
were allotransplanted but only six of ten planned for from the International Bone Marrow Transplant Registry
autologous bone marrow transplantation (ABMT). The
(IBMTR) with a probability of six years leukaemia-free
study illustrates the problems of patient selection during
survival of more than 50% (87). In this retrospective
the course of the disease, but points to a possibility of
analysis of data from the IBMTR between 1987 to 1994,
comparing patients with suitable donors with those with-
1 483 AML patients with HLA-identical sibling transplan-
out donors on the basis of intention-to-treat principles.
tation in rst CR are included.

Some prospective trials have compared the results in
A French group has reported a ten years experience with
patients with an available allogeneic donor with those in
allotransplantation in young (B 45 years) AML patients
patients without a donor according to intention-to-treat
(88). One hundred and thirteen allogeneically transplanted
principles. In a French study 78 patients younger than 40
patients were compared with 173 patients with autologous
years with a suitable allogeneic donor were included (85).
transplantation or with high dose chemotherapy as post-
Fifty-eight patients (74%) achieved a CR, of whom 31
consolidation. All patients got identical induction and rst
patients without a donor were treated with an intensive
consolidation chemotherapy. The data were analyzed ac-
post-induction chemotherapy combination. The cumula-
cording to intention-to-treat principles. The three-year
tive risk of relapse was 67% in this group compared with
leukaemia-free survival and total survival were higher with
43% in 27 patients with an HLA-identical sibling donor
(p¾ 0.01). However, the overall survival rates were not allogeneic transplantation (58% and 64%) compared with
statistically different between the groups. Thus, when com- other post-consolidation therapy (37% and 45%) (pB
paring patients in CR, the availability of an HLA-identical 0.006 and pB 0.07, respectively). The French group has
sibling donor was not associated with a better survival also performed a prospective comparison of allogeneic
rate. This might be explained by the fact that some pa- bone marrow transplantation (alloBMT), autologous stem
tients with a donor did not receive BMT. Moreover, the cell transplantation (ASCT) and chemotherapy (Chemo) in
salvage treatment of patients who relapsed after intensive 204 adult patients with de novo AML (89). Totally 162
consolidation chemotherapy was ef cient. patients (79%) achieved a complete remission (CR) on
A summary of the results from 14 prospective trials induction chemotherapy. Of the 135 patients who were still
comparing transplantation in patients with matched sib- in CR after consolidation, 96 patients were less than 46
lings to chemotherapy in patients lacking family member years of age: 36 patients had an HLA-identical sibling
donors has been published (86). The cure rates with mar- donor and were allocated for alloBMT (group I); they
row transplantation ranged from 40 to 64%, compared were compared with the 60 other patients who did not
with 19 to 24% with chemotherapy. have an HLA-identical sibling donor and were treated
In the eld of transplantation most evidence is derived with either ASCT or chemotherapy (group II). The three-
from retrospective and prospective single centre studies year disease-free survival was higher for group I patients
and register data. Retrospective analyses of data from (67 9 16%) than for the 60 group II patients (42 9 13%)
1696 patients with AML transplanted in Europe from 1987 (pB 0.05). The actuarial risk of relapse at three years
to 1992 and reported to the acute leukaemia EBMT reg- was signi cantly lower for group I patients (24 9 15%)
istry were published by Gorin in 1996 (68). Groups of than for the other 60 group II patients (569 13%;
p B 0.009). The 99 patients who did not ful ll the inclu- geneic BMT was performed in rst CR in only 180 out of
sion criteria for alloBMT were given intensive chemother- 295 patients with a donor (61%), but another 38 patients
apy including high-dose Ara-C. Patients still in CR were allografted later ( ve in resistant disease, 14 in re-
(n¾ 77) were randomized for either ASCT (n ¾ 39) or lapse and 19 in second CR). In this study the patients
Chemo (n¾ 38). No detectable statistical difference be- were strati ed for risk factors according to the number of
tween ASCT and Chemo was seen for either disease-free courses to achieve CR, white blood cell counts at diagno-
survival, risk of relapse or survival. sis and the FAB-classi cation. Cytogenetics was not taken
Several prospective randomised studies have been per- into account. The disease-free survival remained signi -
formed comparing conventional consolidation and au- cantly longer for patients with a donor also after strati -
tologous stem cell transplantation (79, 90 –92). In some of cation for prognosis.
these studies the value of allogeneic transplantation is also The GOELAM study (91) compared three intensive
assessed (90 –92). In the EORTC-GIMEMA AML-8A consolidation strategies in patients with AML in CR 1;
trial (92), autologous transplantation led to better results allogeneic or autologous BMT or intensive consolidation
than conventional treatment in young patients in rst chemotherapy (ICC). Patients aged 15 to 50 years with de
remission. In this study, patients less than 60 years were novo AML received an induction treatment with seven
randomised after attaining a CR after induction with a days of cytarabine and either idarubicin or rubidazone.
‘10 »3’ regimen (Ara-C, etoposide and either idarubicin, After attaining a CR, patients up to the age of 40 with a
daunorubicin or mitoxantrone). All patients got one con- HLA-identical sibling underwent an allogeneic BMT. All
solidation with intermediate dose Ara-C » the same another patients received one consolidation course (high-
thracyclin as in the induction period. Totally, 254 patients dose cytarabine and the same anthracycline as for induc-
without a histocompatible sibling donor were randomised
tion) and were then randomly assigned to either a second
to treatment with high-dose therapy and autologous stem
course of ICC with amsacrine and etoposide or a combi-
cells or a second cycle of consolidation with high dose
nation of busulfan and cyclosphosphamide followed by
Ara-C. One hundred and forty-four out of 168 patients
unpurged ABMT. Of 517 eligible patients, 367 had a CR
with an identical sibling donor got an allogeneic trans-
and were planned for allogeneic transplantation or the

plant. The disease-free survival after transplantation in
other intensive postremission treatment. However, only
rst CR was longer in patients with a HLA compatible
219 (60%) actually received the planned intensive postrem-
sibling; 55% at four years, compared with 48% in the
ission treatment (73 allogeneic BMT, 75 autologous BMT,
autologous group and 30% in the chemotherapy treated
and 71 ICC). With a median follow-up of 62 months, the
patients. The difference in disease-free survival between
ve-year overall survival was 41% for all the 517 eligible
patients treated with autologous transplantation and
patients and the ve-year disease-free survival for the 367
chemotherapy was marginally statistically signi cant (p ¾
patients in CR was 40%. The type of post-remission ther-
0.05). However, the survival at 4 years after CR did not
apy did not have any impact on the outcome, with no
show any difference; 59% and 56%, in patients with allo-
statistically signi cant difference in ve-year DFS and OS
geneic and autologous grafts, respectively. Nor did the
survival of patients in the chemotherapy arm differ, since between the 88 patients for whom an allogeneic BMT was
more patients with relapse in this group had a response to scheduled (44% and 53%, respectively) and the 134 pa-
subsequent autologous transplantation. The results from tients without an HLA-identical sibling (38% and 53%,
the AML 8 study were reanalysed on the basis of inten- respectively). Similarly, there was no difference in the
tion-to-treat principles in a later report (93). Patients B 46 outcome between autologous BMT and ICC. The ve-
years of age, who were alive at eight weeks or more from year DFS was 44% for the 86 patients randomly assigned
diagnosis, were evaluated with the intention to determine to autologous BMT and 40% for the 78 patients assigned
whether patients with a HLA-identical sibling had a sur- to ICC (p¾ 0.41). The ve-year overall survival was simi-
vival advantage over patients without a donor. In this lar in the two groups (50% vs 54%; p¾ 0.72). The median
analysis the number of patients was higher than in the duration of hospitalization and thrombocytopenia was
original report by Zittoun et al. (92), as patients were longer after ABMT (39 v 32 days; p¾ 0.006, and 110 vs
followed from an early point in time and also patients 19 days; p¾ 0.0001, respectively). In summary, after one
treated with salvage therapy were included. The disease- consolidation course with ICC, a second course of
free survival from the time point of CR for the 295 chemotherapy is less myelotoxic than unpurged ABMT,
patients with a donor was signi cantly longer than for 377 but yields comparable disease-free and overall survival
patients without a donor (46 vs 33% at 6 years; p¾ 0.01) rates.
due to a lower relapse rate (42% vs 63%; pB 0.001). The The US Intergroup study randomised 233 patients be-
corresponding gures for overall survival from diagnosis tween the age of 16 –55 years in CR after standard induc-
were 48% vs 40% (p ¾ 0.22). However, a difference in tion chemotherapy and one consolidation (including two
survival semed to occur after the second year. The allo- days of idarubicin and ve days of cytarabine) to a single
course of high dose cytarabine or transplantation of au- In the MRC AML 10 trial described above the indepen-
tologous marrow treated with perfosfamide (90). One hun- dent prognostic signi cance of pretreatment cytogenetics
dred and thirteen patients with a histocompatible sibling was determined (3) and their impact on the outcome of
were offered allogeneic transplantation. In an intention to subsequent transplantation procedures was also evaluated.
treat analysis no signi cant differences were found in AML associated with t(8;21), t(15;17) or inv (21) predicted
disease-free survival between patients receiving chemother- a relatively favourable outcome compared with patients
apy and those undergoing autologous transplantation (me- with a normal karyotype. The presence of a complex
dian 18 and 14 months; ns). These results did not differ karyotype, ¼ 5, del(5q), ¼ 7, or abnormalities of 3q
signi cantly from the median disease-free survival of 32 de ned a group with relatively poor prognosis. The re-
months with allogeneic transplantation. The median fol- maining group of patients including those with 11q23
abnormalities, »8, »21, »22, del(9q), del(7q) were
low-up in this study was ve years. Survival at four years
found to have an intermediate prognosis. Subgroup analy-
after CR seemed to be better after chemotherapy than
sis demonstrated that the three cytogenetically de ned
after autologous marrow transplantation (52% vs 43%;
prognostic groups were found to be a key determinant of
p ¾ 0.05). A marginal survival advantage was seen with
outcome from autologous or allogeneic transplantation in
chemotherapy also when compared with allogeneic trans-
rst CR.
plantation (46% survival at four years; p ¾ 0.04). There
The importance of diagnostic cytogenetics as an inde-
were no signi cant differences in survival between patients

pendent prognostic factor in AML, led to a strati ed
receiving allogeneic marrow and those receiving au- treatment approach, which is adopted in the current MRC
tologous marrow. AML 12 trial. Good risk patients, any patient with fa-
The three above trials are not fully comparable. A more vourable karyotypic abnormalities irrespective of response
effective regimen (high dose of cytarabine) was used in the status after the rst induction course, are not recom-
chemotherapy comparison group in the two studies with- mended for either allogeneic or autologous transplanta-
out advantage in disease-free survival for autotransplanted tion. For standard risk patients, patients with neither
patients, the GOELAM (91) and the U.S. intergroup trial favourable nor adverse karyotype abnormalities and with
(90), than in the EORTC-GIMEMA AML-8A (92). In the less than 15% blasts in the marrow after the rst course,
MRC AML 10 trial a transplantation was added to inten- the value of giving additional chemotherapy before the
sive chemotherapy (79) and thus three options were com- transplantation is investigated. The BMT is randomly
pared; allogeneic or autologus transplantation or no performed after four or ve courses of consolidation; the
further therapy. After four intensive chemotherapy BMT is allogeneic if a sibling donor is available, otherwise
courses, 381 patients in rst CR without an allogeneic autologous. In the study, poor risk patients are excluded
donor were randomised to auto-BMT or no further ther- (patients with \ 15% blasts in the bone marrow after
apy. The intensive chemotherapy regimens were to ensure induction course 1 and without a favourable karyotype).
that all patients got the best available chemotherapy. The These patients proceed to a protocol evaluating a further
study also took into account the in uence of risk category increase in dosage of drugs in combination with cy-
(see below). There was a highly statistically signi cant closporin to overcome multidrug resistence to improve
reduction in relapse risk in the autologous BMT group remission rates.
compared with chemotherapy only (37 vs 58%; p ¾ Both autologous and allogeneic transplantation thus
seems to have a better antileukemic effect with lower
0.0007), and a signi cant advantage in disease-free survival
relapse rate than chemotherapy only. However, due to an
at seven years (53 vs 40%; p ¾ 0.04). The proportional
excess of TRM any overall survival bene t may emerge
reduction in relapse was similar in all risk groups and age
only after longer follow-up. A problem with studies with
groups, although the absolute bene ts varied between the
high-dose consolidation regimens is that only a fraction of
groups. There were excess deaths in remission during the
patients will indeed proceed to their planned therapy. In
pancytopenic phase with autologous transplantation (12 vs
most of the above mentioned studies only about half of the
4%; p¾ 0.008), resulting in equal survival in the rst two patients potentially eligible for autologous transplantation
years. However, a late survival advantage emerged in the were in fact randomised and not all randomised patients
auto-BMT group once patients had reached two years got the transplant. Also patients with an allogeneic donor
(p¾ 0.006). At seven years, 57% of the autotransplanted do not always proceed to transplantation. This fact is
patients were survivors compared with 45% in the taken into account in the intention-to-treat-analyses, but
chemotherapy group. On a donor vs no-donor analysis means that the high-dose treatment is applicable only to a
there was no advantage for the group with an allogeneic proportion of AML patients.
donor regarding overall survival. Patients in rst CR with A minimal or reduced myeloablative conditioning
low risk factors did not bene t from allogeneic transplan- (‘mini-transplantation’ or non-myeloablative stem cell
tation because of a high risk of TRM and the possibility of transplantation ‘NST’) has been introduced to try to
salvage transplantation in second remission. reduce the TRM and morbidity of allogeneic transplanta-
tion. The conditioning chemotherapy is used for destruc- practice for consolidation in rst remission for young
tion of both tumour cells and immunocompetent recipient patients with an HLA-matched sibling. It is however not
cells and the GVL effect is important for cure of the known which patients will really bene t from transplan-
disease. Evidence for the GVL effect has been obtained tation as no truly randomised comparison of allogeneic
from retrospective studies analyzing the relapse incidence vs autologous transplantation or chemotherapy has been
in patients after syngenic transplantation and in patients performed. Patients with and without an HLA-identical
with graft-vs-host disease (72), but also by the successful sibling have been compared on the basis of intention-to-
use of alloreactive donor lymphocyte infusions (DLI) for treat principles (‘genetic randomisation’). The disease-free
the treatment of relapse after allogeneic transplants (94). survival seems to be prolonged in the donor group, due
NST has been shown to induce a host-vs-graft tolerance to a lower relapse rate with allogeneic transplantation.
and an immune graft-vs-leukaemia effect with clinical tu- However, the higher procedure-related mortality makes
mour effects and with a low TRM (95, 96). The minimal the effect on total survival uncertain. According to results
conditioning has made it possible to treat also patients too from retrospective studies, a delay of transplantation to
vulnerable for conventional transplantation. Improve- CR 2 has been suggested in patients with initially good
ments of quality of life are awaited. However, the long- predictors for prognosis.
term outcome of this concept of cellular immunotherapy is Randomised trials indicate a trend for improved disease-
unknown and needs further evaluation in controlled clini- free survival for autologous transplantation vs conven-
cal studies. tional consolidation without clear effects on survival. In
Patients with relapsed AML after a short rst remission one study, in which an autograft was added to four courses
cannot be cured with chemotherapy in conventional doses of intensive therapy, there was also a late survival
(97). However, 25 –35% of patients who achieve a second advantage.
CR followed by transplantation can be cured (67– 69). Thus, the role for intensi ed post-remission treatment in
According to EBMT register data, leukaemia-free survival rst complete remission with high-dose chemotherapy
at ve years seemed to be longer in 624 patients treated followed by allogeneic or autologous marrow or stem cell
with an allogeneic transplantation in second CR (41%)
transplantation requires further studies.

compared with those 708 with autologous transplantation For patients in their second complete remission as well
(33%) (83). The risk of relapse was lower in allo-engrafted as for those refractory to conventional chemotherapy,
patients; 36% compared with 57% in autografted patients. uncontrolled data indicate that allogeneic transplantation
Another retrospective analysis of therapy for patients in might be bene cial. The role of high dose chemotherapy
second CR compared the results of allogeneic transplanta- and transplantation in these settings needs further evalu-
tion in 257 patients reported to the IBMTR with 244 ation.
patients who were treated with chemotherapy according to Allogeneic stem cell transplantation after minimal or
speci c protocols in three group trials (69). The results reduced myeloablative conditioning (‘mini-transplanta-
showed that transplantation was associated with longer tion’ or non-myeloablative stem cell transplantation
disease-free survival than chemotherapy. ‘NST’) is a new concept, which induces a host-vs-graft
The ef cacy of ABMT has been analyzed according to tolerance and an immune graft-vs-leukaemia effect with
the duration of rst remission in a few studies. Petersen et a low procedure related mortality. The long-term effects
al. (98) reported a two-year disease-free survival of 47% in of this type of immunotherapy are unknown and evalu-
patients with a long rst CR compared with 23% in
ation in controlled clinical studies is required.
patients with shorter rst CR. On the other hand, Chopra
et al. (99) found no correlation between disease-free sur- Allogeneic and autologous heamatopetic stem cell
vival after ABMT with the duration of rst CR. In all transplantation:
transplantation studies, it is likely that poor prognosis
patients are excluded, and consequently the results cannot
easily be compared with the results of conventionally Scienti c evidence*
dosed chemotherapy. 1 ¾ High 2¾ Moderate 3¾ Low Total
Allogeneic transplantation seems to be the only form of Number of studies:number of patients
therapy for curing patients with ‘‘refractory’’ AML not
responding to conventional induction therapy. Some stud- M – – – –
C 1:381 3:651 1:78 5:1 110
ies have reported long-term survival in 15 to 20% of these
P 3:2 247 3:370 5:348 11 :2 965
patients (100, 101). No randomised studies have been R 4:5 793 3:935 5:4 007 12 :10 735
performed regarding the effect of transplantation in refrac- L 2 – – 2
tory patients. O 4 – – 4
The literature shows that: Total 14 :8 421 9:1 956 11 :4 433 34 :14 810
Allogeneic bone marrow transplantation is an established *For abbreviations, see above.
TREATMENT OF ELDERLY PATIENTS cline dose was reduced and a small proportion (5%) was
AML is primarily a disease of the elderly and more than untreated and still survived for more than one year.
50% are over 60 years of age at diagnosis (102, 103). In another study, data indicated that older patients
Treatment of elderly patients with AML is still controver- respond to intensive chemotherapy in a similar manner to
sial and their survival has not improved as it has for younger patients (117). One hundred and seven previously
younger patients. Older patients show a marked suscepti- untreated patients with AML with an age ranging from 15
bility to treatment toxicity, which is why the poorer results to 82 years received a 70-day remission induction regimen
may be due to host-related factors, but also to intrinsic consisting of daunorubicin, cytarabine and thioguanine
differences in the biology of the leukaemia (104). Elderly (TAD). Identical CR rates of 65% were observed for 33
patients often have unfavourable chromosomal changes patients 60 years of age and older and for 74 patients aged
and other characteristics of the leukemic cells, which are 15 – 59 years. Median remission duration and survival were
associated with resistence to chemotherapy (105). A 14 and 22 months, respectively, for patients 60 years and
myelodysplastic syndrome often precedes the AML (106) older, and 16 and 22 months, respectively, for patients
and about 10 – 20% of primary AML include three lineage 15 – 59 years. These differences were not statistically
myelodysplastic features at diagnosis (107). According to signi cant.
Leith et al. (105), AML in elderly patients may be viewed In an EORTC trial including newly diagnosed AML
as a secondary leukaemia. Both myelodysplasia and sec- patients above the age of 65 two different therapeutic
ondary AML are associated with low response rates to strategies were compared (112). Patients randomised to
chemotherapy, which may explain the poor prognosis for arm A were given immediate intensive induction therapy
some elderly patients. On the other hand, some elderly while in arm B a wait-and-see policy was used, including
patients show a ‘smouldering’ AML (108) or a hypocellu- supportive care with mild cytoreductive chemotherapy
lar AML (109), which are slowly progressing for a period only for relief of symptoms. Thirty-one patients on arm A
of months to years also without therapy. received daunorubicin, vincristine and cytarabine in one or
two courses for remission induction followed by one addi-
tional cycle of consolidation for patients in CR. Among

Induction treatment the 29 patients on arm B, hydroxyurea (6 ½ 0.5 g:d day 1
and 4) and cytarabine (200 mg:m2 day 1– 4 sc) were
A combination of ara-C with an anthracycline is the
instituted at a median of nine days after diagnosis due to
treatment of choice for many elderly patients with good
leukaemia-associated symptoms. Overall survival was sig-
performance status and absence of concomitant diseases.
ni cantly longer for patients treated with intensive
In most early large studies and co-operative group trials
chemotherapy (median survival 21 vs 11 weeks; p¾ 0.02
only 40 –50% of elderly patients enter CR, and 30 –50% of
and a projected survival of 13 vs 0% at 2.5 years). For 18
the patients die during the induction period (15, 38, 110).
patients in CR group A the two-year survival was 17%.
More recent trials seem to give a better short-term
Moreover, the median percentage of days spent in hospital
outcome with CR rates of 50 –70% and 10 – 25% early
mortality (21, 111, 112). Better supportive care or im- did not differ between the two arms. Thus, supportive care
provement in chemotherapy may explain the results, but with a wait-and-see strategy seemed inferior. This conclu-
also a more adapted selection of patients. For older pa- sion is hampered by the selection bias of patients; probably
tients with poor performance status or high-risk concomi- only a minority of the total patient population was in-
tant diseases, less aggressive therapies are used. Low-dose cluded in the study. The results could be compared with a
ara-C could yield a hematological improvement and in population-based study on AML patients in northern Eng-
some cases also a remission (113 –115). In most studies land (103). Out of 200 patients older than 55 years, 84
with elderly patients, the degree of selection bias is high. A received conventional induction chemotherapy with a me-
retrospective analysis on 235 patients with AML aged 60 dian survival of ve months compared with two months
years or more, who were consecutively admitted to a single for all patients.
hematological department during a three-year period, was Oral induction with etoposide, 6-thioguanine and idaru-
presented by Baudard et al. (116). Forty-six of the patients bicin was compared with a short TAD(1 »5) in a Finnish
received conventional induction chemotherapy but were randomised study (118). In the oral regimen etoposide was
not included in EORTC co-operative clinical trials (indi- given 80 mg:m2 , thioguanine 100 mg:m2 tid on d 1 –5 and
cating a selection bias). Compared with treatment results idarubicin 15 mg:m2 d 1 –3. TAD consisted of thioguanine
in patients less than 60 years, CR rate was lower (33% vs and iv cytarabine both in a dose of 100 mg:m2 tid d 1– 5
65%; pB 0.0001) with a marked drop in patients older and daunorubicin 60 mg:m2 d 5. The oral regimen resulted
than 70. Moreover, induction death rate was higher (21% in a higher remission rate (60% vs 23%; p ¾ 0.007) and
vs 12%; p ¾ 0.04). Some of the elderly patients were sug- longer median survival (9.9 months vs 3.7 months; p¾
gested to have been ‘undertreated’ because the anthracy- 0.04) but with no more toxicity. An ongoing AML 13
EORTC trial in older patients compares intravenous and vival is 8% in each arm and overall survival estimates 6 vs
oral regimens combining idarubicin and etoposide with 9% (ns). Poor performance status at diagnosis, high WBC
ara-C. count, older age, secondary AML, and presence of cytoge-
The doses of daunorubicin administered to elderly pa- netic abnormalities had an adverse impact on survival.
tients mostly range from 30 –60 mg:m2 per day for three Among complete responders, 74 patients were assigned to
days. The German AML Cooperative Group compared ara-C and 73 to no further therapy. Actuarial DFS was
daunorubicin at a dose of 30 mg:m2 or 60 mg:m2 for three signi cantly longer (p ¾ 0.006) for ara-C-treated patients
consecutive days in patients \ 60 years in combination (13%) at ve-years compared with non-treated patients
with standard dose ara-C for seven days with a resulting (7%), but overall survival was similar (p¾ 0.3); 18% vs
higher CR rate (52% vs 45%; p ¾ 0.026) and a lower early 15%. In summary, in previously untreated elderly patients
death rate 20% vs 31%; p ¾ 0.031) in the high-dose group with AML, mitoxantrone induction therapy produces a
(119). In a subgroup of patients \ 65 years of age (n ¾ slightly better CR rate than a daunorubicin-containing
210) the CR rate was 52% vs 32% (p¾ 0.006) and their regimen, but has no clear effect on remission duration and
overall survival was 14 vs 5% (p¾ 0.002). survival. Ara-C in maintenance may prolong disease-free
A randomised trial compared high-vs standard-dose mi- survival, but not overall survival.
toxantrone with ara-C in patients \ 60 years of age as Different factors predictive for prognosis were evaluated
induction without consolidation (120). The dose of mitox- also in the trial of Bow et al. (121). A CD34 stem cell
antrone was either 80 mg:m2 on day 2 or 12 mg:m2 on phenotype, poor performance status and complex cytoge-
days 1 –3 in combination with ara-C 3 g:m2 on days 1– 5. netic aberrations were all independent covariates of failure
No differences in remission rates and survivals were seen, to achieve remission. A prognostic factor analysis of 694
patients older than 60 years entered into the MRC 11 trial
and both regimens gave similar results compared with
identi ed good risk patients who bene t from intensive
standard induction with repetitive consolidation.
therapy and also extremely poor risk patients who do not.
Mitoxantrone was also used in a non-randomised study
In a retrospective study, 131 patients with AML not
in combination with etoposide (10 mg:m2 and 100 mg:m2 ,
eligible for standard chemotherapy with a median age of
respectively both during ve days) as induction for un-

66 years (15 –84) were treated with low dose cytarabine
treated AML patients aged 60 – 80 (121). The safety and
(122). Totally 22 (17%) achieved CR, while 18 died during
ef cacy and impact on quality of life of this non-cytara-
induction. Eighty-seven of the patients were previously
bine containing remission induction therapy were evalu-
untreated and median survival for these patients was 22.5
ated. Patients in CR received a single course of cytarabine
weeks. The prognostic value of some initial disease
0.5 mg:m2 every 12 hours on days 1– 6. A total of 37 (55%)
parameters was evaluated. Low dose cytarabine is effective
of 67 eligible patients achieved remission and the median
for some patients but not for those with a hypercellular
disease-free and overall survival were eight and nine
marrow. Detourmignies et al. (113) treated 77 elderly
months, respectively. Moreover, patients in CR and PR
patients with AML with cytarabine 10 mg:m2 :12 h over
exhibited signi cantly improved quality of life (global 21 days for one or two courses. Thirteen (17%) patients
FLIC scores). achieved CR, 16 (21%) PR and 28 (35%) had resistant
Optimization of both remission-induction and post-re- leukaemia. Twenty patients (26%) died early or during
mission therapy was the subject of a large randomised hypoplasia. Most (86%) patients had severe pancytopenia
intergroup study in 489 patients older than 60 years (30). and 58% were hospitalized. Overall median survival was
Remission-induction chemotherapy was compared be- only three months, but patients in CR had a median
tween daunorubicin 30 mg:m2 on days 1, 2, and 3 vs disease-free survival of nine months, in ve lasting more
mitoxantrone 8 mg:m2 on days 1, 2, and 3, both plus than one year and in two longer than four years. In this
ara-C 100 mg:m2 on days 1– 7. Following CR, patients study, patients with complex cytogenetic changes or rear-
received one additional cycle of daunorubicin or mitox- rangements of chromosomes ve or seven had a very low
antrone and were then eligible for a second randomisation response rate and a median survival of only 1.5 months.
between eight cycles of low-dose (LD)-ara-C 10 mg:m2
subcutaneously every 12 hours for 12 days every six weeks
or no further treatment. A total of 242 patients were Post -remission treatment
randomised to daunorubicin and 247 to mitoxantrone. Totally less than 60% of elderly patients achieving CR
Median age of both groups was 68 years. Secondary AML undergo post-remission treatment (123). Moreover, the
was documented in 26 and 25% in either arm. The proba- post-remission treatment is frequently interrupted because
bility of attaining CR was 47% with mitoxantrone and of poor tolerance. High-dose ara-C consolidation did not
38% with daunorubicin (p¾ 0.07) and with chemotherapy demonstrate signi cant bene t compared with less aggres-
resistence in 32% and 47%, respectively (p¾ 0.001). With a sive regimens (44). Intensive consolidation may, however,
median follow-up of six years, ve-year disease-free sur- be of bene t for selected elderly patients (124).
In some elderly patients unable to tolerate consolidation is still questionable (126). A prospective randomised
tion, a continuous maintenance chemotherapy based on multicentre trial compared the effects of granulocyte-
prolonged administration of ara-C and 6-thioguanine may macrophage colony-stimulating factor (GM-CSF) as an
delay leukaemia relapse and prolong survival duration (48, adjunct to intensive chemotherapy in patients 61 years
111, 125). and older with untreated newly diagnosed AML (127).
In summary, therapeutic attitudes in patients over 60 Patients received either daunorubicin-ara-C with or with-
years may vary from conventional chemotherapy to out GM-CSF. Based on the rationale that GM-CSF might
palliative treatment or supportive care. It is possible to sensitize the leukemic cells to the cytotoxicity of
achieve prolonged survival for a subgroup of elderly pa- chemotherapy as well as enhance white blood cell regener-
tients with conventional chemotherapy, but the treatment ation, GM-CSF was given during chemotherapy as well as
is too toxic for many of them. A risk strati cation after chemotherapy. Patients were treated with one, and in
of AML patients as to performance status, chromo- cases of a PR, with two remission induction cycles. When
somal aberrations and other predictors of prognosis is of CR was attained they received one additional cycle of
special importance for the choice of therapy in older consolidation therapy. Of 318 evaluable patients with a
patients. median age of 68 years, 157 were randomised to receive
The literature shows that: GM-CSF and 161 were assigned to control therapy. The
A combination of ara-C with an anthracycline in con- effect of GM-CSF was evaluated according to intention-
ventional doses, is the treatment of choice to prolong to-treat.
survival for most elderly patients with a good perfor- CR rates did not differ between groups (56 and 55%;
mance status and absence of concomitant diseases. p ¾ 1.0). Recovery of neutrophils was signi cantly faster in
Palliative treatment including oral cytotoxic drugs or GM-CSF-treated patients. The median time of recovery of
low dose ara-C subcutaneously is a treatment option for neutrophils towards 0.5 ½ 109 :L was 23 days in the GM-
patients with a poor performance status and concomi- CSF group vs 25 days in the control group (p¾ 0.0002)
tant diseases and for old patients with a hypocellular with the percentages of patients who recovered being 81%
marrow. and 71%, respectively. The probabilities of survival at two

There is still a lack of trials comparing supportive years were 22% in both groups (p ¾ 0.55). Disease-free
therapy only and different low-dose therapy concepts. survivals at two years were 15% and 19% (p¾ 0.69). The
No studies have prospectively analyzed the effect of number of nights spent in the hospital, the number of
therapies in different prognostic groups. transfusions, and the frequencies and types of hemor-
rhages and infections did not differ. Apart from a faster
neutrophil recovery, GM-CSF during and after induction
Treatment of elderly patients:
chemotherapy did not improve the outcome of elderly
patients with AML.
Scienti c evidence* The value of GM-CSF was assessed in another group of
patients 60 years of age or older (median age, 69 years),
1 ¾High 2 ¾ Moderate 3 ¾Low Total
Number of studies:number of patients
with newly diagnosed primary AML (128). The patients
were randomized to receive placebo or GM-CSF ( ve
C 3:2 309 6:1 107 4:473 13:3 889 micrograms per kilogram of body weight per day intra-
P 1:1 088 6:488 – 7:1 576 venously over a period of six hours) in a double-blind
R 1:235 3:356 – 4:591
manner. Growth factor therapy started the day after the
O 2 – – 2
L 3 – – 3 completion of three days of daunorubicin (45 mg:m2 :day)
Total 10 :3 632 15: 1 951 4:473 29:6 056 and seven days of cytarabine (200 mg:m2 :day by continu-
ous intravenous infusion). If leukemia cells persisted in the
marrow three weeks after the initiation of chemotherapy,
*For abbreviations, see above. further daunorubicin (two days) and cytarabine ( ve days)
were administered. GM-CSF or placebo was given daily
until the neutrophil count was at least 1 000 per cubic
millimeter, there was evidence of the regrowth of leukemia,
COLONY-STIMULATING FACTORS or severe toxic effects attributable to the study infusion
The use of recombinant colony-stimulating factors (CSFs) occurred. Patients who had a complete remission were
has been studied in AML because of the possibility of randomly assigned to receive one of two intensi cation
accelerating the recovery of normal hematopoiesis. This regimens. Totally, 193 patients were randomly assigned to
might be especially important for older patients. Despite receive GM-CSF and 195 to placebo. The rate of complete
several randomised studies, the bene t of CSF administra- remission was 51 percent among patients assigned to GM-
CSF and 54 percent among those assigned to receive Scienti c evidence*

placebo (p¾ 0.61). The reasons for failure (early death, 1 ¾High 2 ¾ Moderate 3 ¾ Low Total
death during marrow hypoplasia, and persistent leukemia), Number of studies:number of patients
the incidence of severe or lethal infection, and the inci-
dence of the regrowth of leukemia were similar in the two M 1:1 897 – – 1:1 897
C 14:10 096 27:5 621 10 :1 676 51 :17 393
groups. The median duration of neutropenia was slightly
P 8:4 891 21:1 553 10 :2 007 39 :8 451
shorter in the patients who received GM-CSF (15 days) R 5:6 028 8:1 744 5:4 044 18 :11 816
than in those who received placebo (17 days), (p ¾ 0.02). L 10 – – 10
The clinical importance of this result was minimal because O 10 – – 10
the growth factor failed to lower the treatment-related Total 48:22 912 56:8 918 25 :7 727 129: 39 557
mortality rate or improve the rate of complete remission.
*For abbreviations, see above.
Heil et al. (129) assessed the safety and ef cacy of
lgrastim (G-CSF) as an adjunct to AML induction and REFERENCES
consolidation therapy in a prospective double-blind, ran-
1. Bloom eld CD, Shuma C, Regal L, et al. Long-term survival
domised, placebo-controlled, multicentre trial. A total of
of patients with acute myeloid leukemia: a third follow-up of
521 consecutive de novo AML patients (302 aged 50 or the Fourth International Workshop on Chromosomes in
more) were randomised to receive lgrastim (5 g:kg:d Leukemia. Cancer 1997; 80 (Suppl): 2191 –8. (L)
subcutaneously) or placebo after standard induction as 2. Bloom eld CD, Lawrence D, Byrd JC, et al. Frequency of
well as after consolidation chemotherapy. The overall CR prolonged remission duration after high-dose cytarabine in-
tensi cation in acute myeloid leukemia varies by cytogenetic
rate was 68%. Median disease-free survival was ten months
subtype. Cancer Res 1998; 58: 4173 – 9. (R2:285)
and median overall survival 13 months with no differences 3. Grimwade D, Walker H, Wheatley K, et al. The importance
between treatment groups. Patients receiving lgrastim of diagnostic cytogenetics on outcome in AML: analysis of
experienced neutrophil recovery ve days earlier after in- 1(612 patients enterred into the MRC AML 10 trial. The
duction than those receiving placebo (pB 0.0001). This Medical Research Council Adult and Children’s Leukaemia
Working Parties. Blood 1998; 92: 2322 – 33. (R1:1612)
was accompanied by reductions in the duration of fever (7
4. Ellison RR, Holland JF, Weil M, et al. Arabinosyl cytosine:

vs 8.5 days; p ¾ 0.009), parenteral antibiotic use (15 vs 18.5 A useful agent in the treatment of acute leukaemia. Blood
days; p¾ 0.0001), and hospitalization (20 vs 25 days; 1968; 32: 507 – 23. (C2:234)
p ¾ 0.0001). Similar reductions were seen after induction 2 5. Plunkett W, Iacoboni S, Keating MJ, et al. Cellular pharma-
cology and optimal theurapeutic concentrations of 1-beta-D-
and the consolidation courses. There was a signi cant
arabinofuranosylcytosine triphosphate in leukemic blasts
reduction in the number of patients requiring systemic during treatment of refractory leukaemia with high dose
antifungal therapy in the lgrastim group during induction 1-beta-D-arabino furanosylcytosine. Scand J Haematol 1986;
treatment (34% vs 43%; p¾ 0.04). 44: 51 – 9. (O)
The literature shows that: 6. Dillman RO, Davis RB, Green M, et al. Comparison of ara-C
at 200 mg:m2 and 100 mg:m2 in induction and maintenance
The use of G-CSF and GM-CSF is safe, but has no therapy for acute myelocytic leukaemia. Proc Am Soc Clin
impact on the prognosis in elderly AML patients. Oncol 1987; 6: Abstr 160. (C3:98)
7. Dillman RO, Davis RB, Green M, et al. A comparative study
The duration of neutropenia and infectious complica- of two different doses of cytarabin for acute myeloid
tions are reduced. leukaemia: A phase III trial of Cancer and Leukaemia Group
B. Blood 1991; 78: 2520 – 6. (C1:326)
8. Bishop JF, Matthews JP, Young GA, et al. A randomised
Colony-stimulating factors: study of high-dose cytarabine in induction in acute myeloid
leukaemia. Blood 1996; 87: 1710 – 7. (C1:301)
9. Weick JK, Kopecky KJ, Appelbaum FR, et al. A randomised
Scienti c evidence* investigation of high-dose vs standard dose cytosine ara-
binoside with daunorubicin in patients with previously un-
1 ¾High 2 ¾ Moderate 3 ¾ Low Total treated acute myeloid leukaemia, a South West Oncology
Number of studies:number of patients Group study. Blood 1996; 88: 2841 – 51. (C1:665)
10. Boiron M, Jaquillat C, Weil M, et al. Daunorubicin in the
C 2:839 1:117 – 3:956 treatment of acute myelocytic leukaemia. Lancet 1969; 1:
Total 2:839 1:117 – 3:956 330 – 3. (P2:64, 48 adults)
11. Weil M, Glidewell OJ, Jaquillat C, et al. Daunorubicin in the
*For abbreviations, see above. therapy of acute granulocytic leukaemia. Cancer Res 1973;
33: 921 – 8. (O)
12. Büchner T, Urbanitz D, Emmerich B, et al. Multicenter study
on intensi ed remission induction therapy for acute myeloid
leukaemia. Leukaemia Res 1982; 6: 827 – 31. (P1:213)
LITERATURE 13. Mayer RJ. Current chemotherapeutic treatment approaches
The articles included in the reference list were classi ed and to the management of untreated adults with de novo acute
graded as follows (number of studies:number of patients): myelogenous leukaemia. Semin Oncol 1987; 14: 384 – 96. (L)
14. Hewlett J, Kopecky KJ, Head D, et al. A prospective bicin in previously untreated adult patients with acute non-
evaluation of the value of bone marrow transplantation and lymphocytic leukaemia. Leukaemia 1990; 4: 177 – 83. (C2:200)
low dose maintenance chemotherapy in the treatment for 30. Lowenberg B, Suciu S, Archimbaud E, et al. Mitoxantrone vs
adult acute myelogenous leukaemia (AML). Leukaemia 1995; daunorubicin in induction — consolidation chemotherapy —
9: 562 – 9. (P2:522, 295 in CR of whom 150 were randomised, the value of low-dose cytarabine for maintenance of remis-
C3:150) sion, and an assessment of prognostic factors in acute myeloid
15. Yates J, Glidwell O, Wiernik P, et al. Cytosine arabinoside leukaemia in the elderly: nal report. European Organization
with daunorubicin or adriamycin for therapy of acute myelo- of the Research and Treatment of Cancer and the Dutch-Bel-
cytoid leukaemia. A CALGB study. Blood 1982; 60: 454 – 62. gian Hemato-Oncology Cooperative Hovon Group. J Clin
(C2:226) Oncol 1998; 16: 872 – 81. (C1:489)
16. Berman E, Raymond W, Daghestani A, et al. 4-Demethoxy 31. Paul C, Tidefelt U, Gahrton G, et al. A randomised compari-
daunorubicin (idarubicin) in combination with 1-beta D-ara- son of doxorubicin and doxorubicin-DNA in the treatment of
binofuranosyl cytosine in the treatment of relapsed or refrac- acute non-lymphoblastic leukaemia. Leuk Lymphoma 1991;
tory acute leukaemia. Cancer Res 1989; 49: 477 – 81. (P2:51)
3: 335 – 64. (C2:121)
17. Daghestani A, Arlin ZA, Leyland-Jones B, et al. Phase I and
32. Bjorkholm M, Gahrton G, Grimfors G, et al. Mitoxantrone,
II cinical and pharmacological study of 4-methoxy daunoru-
etoposide and ara-C vs doxorubicin-DNA, ara-C, thogua-
bicin (idarubicin) in adult patients with acute leukaemia.
nine, vincristine and prednisolone in the treatment of patients
Cancer Res 1985; 45: 1408 – 15. (P2:28)
with acute myelocytic leuekmia. A randomised comparison.
18. Harrousseau JL, Hurteloup P, Reiffers J, et al. Idarubicin in
Eur J Hematol 1995; 55: 19 – 23. (C3:86)
the treatment of refractory acute myeloid leukaemia. Cancer
Treatm Rep 1987; 71: 991 – 2. (P2:42) 33. Rivera G, Avery T, Roberts D. Response to L1210 to
19. Berman E, Heller G, Santorsa J, et al. Results of a ran- combination of cytosinarabinoside and VM-26 or VP16-213.
domised trial comparing idarubicin and cytosine arabinoside Eur J Cancer 1975; 11: 639 –47. (O)
with daunorubicin and cytosine arabinoside in adult patients 34. Bishop JF, Lowenthal RM, Joshua D, et al. Etoposide in
with newly diagnosed acute myelogenous leukaemia. Blood acute nonlymphocytic leukaemia. Australian leukaemia
1991; 77: 1666 – 74. (C2: 120) group. Blood 1990; 75: 27 – 32. (C1:264)
20. Mandelli F, Petti MC, Ardia A, et al. A randomised clinical 35. Hann IM, Stevens RF, Goldstone AH, et al. Randomised
trial comparing idarubicin and cytarabine to daunorubicin comparison of DAT vs ADE as induction chemotherapy in
and cytarabine in the treatment of acute non-lymphoid children and younger adults with acute myeloid leukaemia:
leukaemia. A multicentric study from the Italian Cooperative results of the Medical Research Council’s 10th AML trial
Group GIMEMA. Eur J Cancer 1991; 27: 750 – 5. (C2:249) (MRC AML 10). Blood 1997; 89: 2311 – 8. (C1:1 857, 1572
21. Reiffers J, Huguet F, Stoppa AM, et al. A prospective adults)
randomised trial of idarubicin vs daunorubicin in combina- 36. Carey RW, Ribas-Momdo M, Ellison RR, et al. Comparative
tion chemotherapy for acute myelogenous leukaemia of the study of cytarabinoside therapy alone and combined with
age group 55 to 75. Leukaemia 1996; 10: 389 – 95. (C2:220) thioguanine, mercaptopurin or daunorubicin in acute myelo-
22. Vogler WR, Velez-Garcia E, Weiner RS, et al. A phase III cytic leukaemia. Cancer 1975; 36: 1560 – 5. (C2:326)
trial comparing idarubicin and daunorubicin in combination 37. Omura GA, Vogler WR, Lefante JR, et al. Treatment of acute
with cytarabine in acute myelogenous leukaemia: A South- myelogenous leukaemia: in uence of three induction regi-
eastern Cancer Study Group. J Clin Oncol 1992; 10: 1103 – 11. mens. Cancer 1982; 49: 1530 – 6. (C2:396)
(C2:218) 38. Preisler H, David RB, Kirshner J, et al. Comparison of 3
23. Wiernik PH, Banks PL, Case DC, Jr, et al. Cytarabine plus remission induction regimens and two post-induction regi-
idarubicin or daunorubicin as induction and consolidation mens for the treatment of acute non-lymphocytic leukaemia.
therapy for previously untreated adult patients with acute Blood 1987; 69: 1441 – 9. (C2:253)
myeloid leukaemia. Blood 1992; 79: 313 – 9. (C2:214) 39. Bishop JF. Intensi ed therapy for acute myeloid leukaemia.
24. Berman E, Wiernik P, Vogler R, et al. Long-term follow-up N Engl J Med 1994; 331: 941 – 2. (C1:264)
of three randomized trials comparing idarubicin and 40. Rees JK, Gray RG, Wheatley K. Dose intensi cation in acute
daunorubicin as induction therapies for patients with un-
myeloid leukaemia: greater effectiveness at lower cost. Princi-
treated acute myeloid leukaemia. Cancer Suppl 1997; 11:
pal report of the Medical Research Council’s AML9 study.
2181 – 5. (C2:552)
MRC Leukaemia in Adults Working Party. Br J Haematol
25. The AML Collaborative Group. A systemic collaborative
1996; 94: 89 –98. (C1:972)
overview of randomised trials comparing idarubicin with
41. Castaigne S, Archimbaud E, Bordessoule D, et al. Sequential
daunorubicin (or other anthracyclines) as induction therapy
induction or double induction chemotherapy increase disease
for acute myeloid leukaemia. Br J Haematol 1998; 103:
100 – 9. (M1:1 897) free survival compared with ‘3 »7’ chemotherapy in less than
26. BjoÈ rkholm M, Bjornsdottir J, Stenke L, et al. Mitoxantrone, 50 years adults with acute myeloid leukaemia (AML). Blood
etoposide and cytarabine in the treatment of acute non- 1996;88(Suppl 1): 291 Abstr 1153. (C3:307)
lymphocytic leukaemia. Med Oncol 1990; 47: 112 – 4. (P2:53) 42. Jehn U. Long-term outcome of post-remission chemotherapy
27. Estey EH, Keating MJ, Mc Credie KB, et al. Phase II trial for adults with acute myeloid leukaemia using different
of mitoxantrone in patients with acute leukaemia in relapse. dose-intensities. Leuk Lymphoma 1994; 15: 99 – 112. (P1:122)
Cancer Treat Rep 1983; 67: 389 – 96. (P3:41) 43. Cassileth PA, Lynch E, Hines JD, et al. Varying intensity of
28. Hiddemann W, Kreutzmann H, Straif K, et al. High dose postremission therapy in acute myeloid leukaemia. Blood
cytosine arabinoside and mitoxantrone: A highly effective 1992; 80: 1924 – 30. (C2:170)
regimen in refractory acute myeloid leukaemia. Blood 1987; 44. Mayer RJ, Davis RB, Schiffer CA, et al. Intensive postremis-
69: 744 – 9. (P3:40) sion chemotherapy in adults with acute myeloid leukaemia.
29. Arlin Z, Case DC, Jr, Moroe J, et al. Randomised multicenter Cancer and Leukaemia Group B. N Engl J Med 1994; 331:
trial of cytosine arabinoside with mitoxantrone or daunoru- 896 – 903. (P1:1088; C1:693)
45. Cassileth PA, Harrington DP, Hines JD, et al. Maintenance geneic bone marrow transplantation for de novo acute
chemotherapy prolongs remission duration in adult acute leukemia. Eur J Haematol 1997; 58: 223 – 40. (P2:118)
non-lymphocytic leukaemia. J Clin Oncol 1988; 6: 583 – 7. 62. Lambertenghi-Deliliers G, Maiolo AT, Annalore C, et al.
(C3:51) Idarubicin in sequential combination with cytosinara-
46. Fopp M, Fey MF, Bacchi M, et al. Post-remission therapy binoside in the treatment of relapsed and refractory acute
of adult acute myeloid leukaemia: one cycle of high-dose vs nonlymphoblastic leukaemia. Eur J Cancer Clin Oncol 1987;
standard-dose cytarabine. Leukaemia Project Group of the 23: 1041 – 5. (P2:16)
Swiss Group for Clinical Cancer Research (SAKK). Ann 63. Clavio M, Carrara P, Miglino M, et al. High ef cacy of
Oncol 1997; 8: 251 – 7. (C2:137) udarabine-containing therapy (FLAG-FLANG) in poor
47. Vogler WR, Weiner RS, Moore JO, et al. Long-term follow- risk acute myeloid leukaemia. Hematol 1996; 81: 513 – 20.
up of a randomised post-induction therapy trial in acute (P2:51)
myelogenous leukaemia. Leukaemia 1995; 9: 456 – 60. (C2: 64. Estey E, Plunkett W, Gandhi V, et al. Fludarabine and
398) arabinosyl cytosine therapy of refractory and relapsed acute
48. Büchner T, Hiddeman W, Wormann H, et al. Late events in myelogenous leukaemia. Leuk Lymphoma 1993; 9: 343 – 50.
AML. Data from long-term observation of patients in two (P2:59)
trials starting in 1981 and 1985. Blood 1997; (Suppl 1): Abstr 65. Krough-Jensen M, Stentoft J. Low-dose cytosine ara-
2247. (C3: 321) binoside in the treatment of relapsed and refractory acute
49. Prentice HG, Macdonald ID, Hamon MD. The role of non-lymphocytic leukaemia. Acta Haematol 1986; 76: 127 –
immunotherapy in the treatment of acute myeloblastic 9. (P3:10)
leukaemia from allogeneic bone marrow transplantation to 66. List AF, Spier C, Greer J, et al. Phase I:II trial of cy-
the application of IL-2. Cancer Treat Res 1993; 64: 121 – 34. closporine as a chemotherapy-resistance modi er in acute
(L) leukaemia. J Clin Oncol 1993; 11: 1652 – 60. (P2:42)
50. Hellstrand K, Mellquist UH, Wallhult E, et al. Histamine 67. Clift RA, Buckner CD, Thomas ED, et al. The treatment of
and interleukin-2 in acute myelogenous leukaemia. Leuk acute non-lymphocytic leukaemia by allogeneic marrow
Lymphoma. 1997; 27: 429 – 38. (P 3:22) transplantation. Bone Marrow Transplant 1987; 2: 243 – 58.
51. Thalhammer F, Geissler K, Kyrle PA, et al. Duration of (R1:363)
second complete remission in patients with acute myeloid 68. Gorin NC, Labopin M, Fouillard L, et al. Retrospective
leukaemia treated wih chemotherapy; a retrospective single- evaluation of autologous bone marrow transplantation vs
center study. Ann Hematol 1996; 72: 216 – 22. (R2:168) allogeneic bone marrow transplantation from an HLA iden-
52. Welborn JL, Lewis JP, Meyers FJ. Impact of reinduction tical related donor in acute myelocytic leukaemia. A study of
regimens on the clinical course of adult acute non- the European Cooperative Group for Blood and Marrow
lymphocytic leukemia. Leukemia 1988; 11: 711 – 6. (L) Transplantation (EBMT). Bone Marrow Transplant 1996;
53. Plunkett W, Iacoboni S, Estey E, et al. Pharmacologically 18: 111 – 7. (R1:1 696; 1 114 adults transplanted in CR1, 288
directed ara-C therapy for refractory leukaemia. Semin in CR2)
Oncol 1985; 2: 20 – 30. (L) 69. Gale RP, Horowitz MM, Rees JK, et al. Chemotherapy vs
54. Rudnick SA, Cadman EC, Capizzi RL, et al. High dose transplants for acute myelogenous leukaemia in second re-
cytosine arabinoside (HIDAC) in refractory acute mission. Leukaemia 1996; 10: 13 – 9. (R2:501)
leukaemia. Cancer 1979; 44: 1189 – 93. (P3: 13) 70. Horowitz MM, Gale RP, Sondel PM, et al. Graft vs
55. Vogler WR, Mc Carley DL, Stagg M, et al. A phase III trial leukaemia reactions after bone marrow transplantations.
of high-dose cytosine arabinoside with or without etoposide Blood 1990; 75: 555 – 62. (O)
in relapsed and refractory acute myelogenous leukaemia. 71. Weiden PL, Flournoy N, Thomas ED, et al. Antileukemic
Leukaemia 1994; 8: 1847 – 53. (C2:131) effect of graft-vs host disease in human recipients of allo-
56. Hiddemann W, Aul C, Maschmeyer G, et al. High dose vs genic-marrow grafts. New Engl J Med 1979; 304: 1529 – 33.
intermediate dose cytosine arabinoside combined with mi- (O)
toxantrone for the treatment of relapsed and refractory acute 72. Ringden O, Hermans J, Labopin M, et al. The highest
myeloid leukaemia. Leuk Lymphoma 1993; 10: 133 – 7. (C2: leukaemia-free survival after allogeneic bone marrow trans-
186) plantation is seen in patients with grade I acute graft-vs-host
57. Archimbaud E, Thomas X, Lebland V, et al. Time sequen- disease. Acute and Chronic Leukaemia Working Parties of
tial chemotherapy for previously treated patients with acute the European Group for Blood and Marrow Transplanta-
myeloid leukemia: long-term follow-up of the etoposide, tion (EBMT). Leuk Lymphoma 1996; 24: 71 – 9. (R1:2122)
mitoxantrone, and cytarabine-86 trial. J Clin Oncol 1995; 13: 73. Beatty PG, Kollman C, Howe CW. Unrelated-donor mar-
11 – 8. (P1:133) row transplants: the experience of the National Marrow
58. Kantarjian HM, Keating MJ, Walters RS, et al. The charac- Donor Program. Clin Transpl 1995; 3: 271 – 7. (L)
teristics and outcome of patients with late relapse acute 74. Schiller G, Feig SA, Territo M, et al. Treatment of advanced
myelogenous leukemia. J Clin Oncol 1988; 6: 232 –8. (R3:58) acute leukaemia with allogeneic bone marrow transplanta-
59. Legha SS, Keating MJ, Mc Credie KB, et al. Evaluation of tion from unrelated donors. Br J Haematol 1994; 88: 72 – 8.
AMSA in previously treated patients with acute leukaemia. (R3:55)
Results of therapy in 109 adults. Blood 1982; 60: 484 – 9. 75. Ringden O, Labopin M, Gluckman E, et al. Donor search or
(P2:109) autografting in patients with acute leukaemia who lack an
60. Hines JD, Oken M, Mazza J, et al. High dose cytosine HLA-identical sibling? A matched-pair analysis. Acute
arabinoside (ara-C) and M-AMSA is effective therapy in Leukaemia Working Party of the European Cooperative
relapsed acute non-lymphocytic leukaemia (ANLL). J Clin Group for Blood and Marrow Transplantation (EBMT) and
Oncol 1984; 2: 545 – 9. (P3:40) the International Marrow Unrelated Search and Transplant
61. Wahlin A, Brinch L, Hornsten P, et al. Outcome of a (IMUST) Study. Bone Marrow Transplant 1997; 19: 963 – 8.
multicenter treatment program including autologous or allo- (R2:191 unrelated, 382 autologous)
76. Jones JR, Vogelsang GB, Hess AD, et al. Induction of graft 92. Zittoun RA, Mandelli F, Willemze R, et al. Autologous or
vs host disease after autologous bone marrow transplanta- allogeneic bone marrow transplantation compared with in-
tion. Lancet 1989; 1: 754 – 7. (O) tensive chemotherapy in acute myelogenous leukaemia. N
77. Talbot DC, Powles RL, Sloane JP, et al. Cyclosporin in- Engl J Med 1995; 332: 217 – 23. (P1:990; 422 CR; 168
duced graft vs host disease after autologous bone marrow allogeneic transplantation; C2:254 autologous transplanta-
transplantation for acute myeloid leukaemia. Bone Marrow tion or chemotherapy)
Transplant 1990; 6: 17 – 20. (P3:9) 93. Keating S, de Witte T, Sucio S, et al. The in uence of
78. Burnett AK, Tansey P, Watkins S, et al. Transplantation of HLA-matched sibling donor availability on treatment out-
unpurged autologous marrow for acute myeloid leukaemia come for patients with AML: an analysis of the AML 8A
in rst remission. Lancet 1984; 2: 1068 – 70. (P3:12) study of the EORTC Leukaemia Cooperative Group and
79. Burnett AK, Goldstone AH, Stevens RMF, et al. Ran- GIMEMA. Br J Haematol 1998; 339: 1344 – 53. (C2:831)
domised comparison of addition of autologous bone-marrow 94. Kolb HJ, Schattenberg A, Goldman JM, et al. Graft-vs-
transplantation to intensive chemotherapy: results of MRC leukaemia effect of donor lymphocyte transfusions in mar-
AML 10 trial for acute myeloid leukaemia in rst remission. row grafted patients. EBMT Chronic Leukaemia Working
The Lancet 1998; 351: 700 – 8. (C1:381) Parties. Blood 1995; 5: 2041 – 50. (P2:135, 23 AML)
80. Imrie K, Dicke KA, Keating A. Autologous bone marrow 95. Giralt S, Estey E, Albitar M, et al. Engraftment of allogeneic
transplantation for acute myeloid leukaemia. Stem Cells hematopoietic progenitor cells with purine analog-containing
1996; 14: 69 – 78. (O) chemotherapy: harnessing graft-vs-leukaemia without
81. Gorin NC, Labopin MM, Meloni G, et al. Autologous bone myeloablative therapy. Blood 1997; 12: 4531 –6. (P3:15)
marrow transplantation for acute myeloblastic leukaemia in
96. Slavin S, Nagler A, Naparstek E, et al. Nonmyeloablative

Europe: further evidence of the role of marrow purging by stem cell transplantation and cell therapy as an alternative to
mafosfamide. Leukaemia 1991; 5: 896 – 904. (R1:773 adults, conventional bone marrow transplantation with lethal cy-
263 autotransplanted) toreduction for the treatment of malignant and nonmalig-
82. Mehta J, Powels R, Singhal S, et al. Autologous bone nant hematologic diseases. Blood 1998; 91: 756 –63. (P3:26)
marrow transplantation for acute myeloid leukaemia in rst 97. Keating MJ, Katarjian H, Smith TL, et al. Response to
remission: identi cation of modi able prognostic factors. salvage therapy and survival after relapse in acute myeloge-
Bone Marrow Transplant 1995; 16: 499 – 506. (R2:74)
nous leukemia. J Clin Oncol 1989; 7: 1071 – 80. (R2:243)
83. Gorin NC, Labopin M. On behalf of the ALWP. Analysis of
98. Petersen FB, Lynch MH, Clift R, et al. Autologous marrow
the acute leukaemia EBMT registry. Bone Marrow Trans-
transplantation for patients with acute myeloid leukemia in
plant 1997; 19: Abstr 307. (R3:3 850 allo »3 366 auto)
untreated rst relapse or in second compete remission. J Clin

84. Ljungman P, de Witte T, Verdonck L, et al. Bone marrow
Oncol 1993; 11: 1353 –60. (R3:47)
transplantation for acute myeloblastic leukaemia: an EBMT
99. Chopra R, Goldstone AH, McMillan AK, et al. Successful
Leukaemia Working Party prospective analysis from HLA-
treatment of acute myeloid leukemia beyond rst remission
typing. Br J Haematol 1993; 84: 61 – 6. (P2:147)
with autologous bone marrow transplantation using busul-
85. Archimbaud E, Thomas X, Michallet M, et al. Prospective
fan: cyclophosphamide and unpurged marrow. The British
genetically randomised comparison between intensive post-
Autograft Group experience. J Clin Oncol 1991; 9: 1840 – 7.
induction therapy and bone marrow transplantation in
(R3:34)
adults with newly diagnosed acute myeloid leukaemia. J Clin
100. Biggs JC, Horowitz MM, Gale RP, et al. Bone marrow
Oncol 1994; 12: 262 – 7. (C3:78)
transplants may cure patients with acute leukaemia never
86. Beatty PG. Bone marrow transplantation for the treatment
of hematologic diseases: status in 1994. Exp Hematol 1995; achieving remission with chemotherapy. Blood 1992; 80:
23: 277 – 88. (L) 1090 – 3. (P2:88)
87. Appelbaum FR. Allogeneic haematopoetic transplantation 101. Forman SJ, Scmidt GM, Nademanee AP, et al. Allogeneic
for acute leukaemia. Sem Oncol 1997; 24: 114 –23. (L) bone marrow transplantation as therapy for primary induc-
88. Reiffers J, Stoppa AM, Hugues F, et al. The role of allo- tion failure for patients with acute leukaemia. J Clin Oncol
geneic transplantation in the younger adults with acute 1991; 9: 1570 – 4. (R3:21)
myeloid leukaemia in rst remission: a ten year experience. 102. Cartwright RA, Staines A. Acute leukaemias. In: Fleming A
Blood 1996; 88 (Suppl 1): Abstr 2721. (P3:286) T, ed. Epidemiology of hematological disease. Part 1, vol. 5.
89. Reiffers J, Stoppa AM, Attal M, et al. Allogeneic vs au- Bailliere’s Clin Hematol, 1992: 1– 26. (L)
tologous stem cell transplantation vs chemotherapy in pa- 103. Taylor PR, Reid MM, Stark AN, et al. De novo acute
tients with acute myeloid leukaemia in rst remission. The myeloid leukaemia in patients over 55-years-old: a popula-
BMGT’87 study. Leukaemia 1996; 10: 1874 – 82. (P2:204, tion based study of incidence, treatment and outcome.
162 CR; C3:77) Northern Region Haematology Group. Leukaemia 1995; 9:
90. Cassileth PA, Harrington DP, Appelbaum FR, et al. 231 – 7. (P2:200)
Chemotherapy compared with autologous or allogenic bone 104. Johnson PR, Hunt LP, Liu Yin JA. Prognostic factors in
marrow transplantation in the management of acute myeloid elderly patients with acute myeloid leukaemia: development
leukaemia in rst remission. New Engl J Med 1998; 339: of a model to predict survival. Br J Haematol 1993; 85:
1649 – 64. (P1:740, induction therapy; 346 in CR, of whom 300 – 6. (P2:104)
113 had an allogeneic transplant and C2:233, post-remission 105. Leith CP, Kopecky KJ, Godwin J, et al. Acute myeloid
therapy) leukaemia in the elderly: Assessment of multidrug resistance
91. Harousseau JL, Cahn J, Pignon B, et al. Comparison of (MDR1) and cytogenetics distinguishes biologic subgroups
autologous bone marrow transplantation and intensive with remarkably distinct responses to standard chemother-
chemotherapy as postremission therapy in adult acute apy. Blood 1997; 89: 3323 – 39. (C2:211)
myeloid leukaemia. The Groupe Ouest Est Leucémies Aiguës 106. Hamblin T. The treatment of acute myeloid leukaemia pre-
Myéloblastiques (GOELAM). Blood 1997; 90: 2978 – 86. (P1: ceded by the myelodysplastic syndrome. Leukaemia Res
517; 367 in CR; C2:164) 1992; 16: 101 – 8. (L)
107. Hast R, Widell S. Dysplastic peripheral blood polymorphs 120. Feldman EJ, Seiter K, Damon L, et al. A randomised trial
link acute myeloblastic leukaemia in elderly to myeloblastic of high-vs standard dose mitoxantrone with cytarabine in
syndromes. Eur J Haematol 1992; 48: 163 – 7. (O) elderly patients with acut myeloid leukaemia. Leukaemia
108. Nussler V, Saver H, Pelka-Fleischerf R, et al. Clinical bio- 1997; 11: 485 – 9. (C3:54)
chemical and cytokinetic parameters for distinguishing 121. Bow EJ, Sutherland JA, Kilpatrick MG, et al. Therapy of
smouldering and rapidly proliferating variants of acute untreated acute myeloid leukaemia in the elderly: remission
leukaemia. Eur J Hematol 1990; 45: 19 – 25. (O) induction using a non-cytarabine-containing regimen of mi-
109. Howe RB, Bloom eld CD, Mc Kennan, et al. Hypocellular toxantrone plus etoposide. J Clin Oncol 1996; 14: 1345 – 52.
acute leukaemia, Am J Med 1982; 72: 391 – 5. (O) (P2:67)
110. Rees JK, Gray RG, Swirsky D, et al. Principal results of the 122. Pascarella A, Marrani C, Leoni F, et al. Low-dose cy-
Medical Research Council’s 8th acute myeloid leukaemia tosinearabinoside in patients with acute myeloid leukaemia
trial. Lancet 1986; 2: 1236 – 41. (C1:1127) not eligible for standard chemotherapy. Leuk Lymphoma
111. Büchner T, Urbanitz D, Hiddemann W, et al. Intensi ed 1995; 18: 465 – 9. (R2:131)
induction and consolidation with or without maintenance 123. Bassan R, Buelli M, Viero P, et al. The management of acute
chemotherapy for acute myeloid leukaemia (AML): two myelogenous leukaemia in the elderly: Ten year experience in
multicenter studies of the German AML Cooperative 118 patients. Hematol Oncol 1992; 10: 25 – 60. (R2:118)
Group. J Clin Oncol 1985; 3: 1583 – 9. (C2:143) 124. Liu Yin JA, Johnson PR, Davies JM, et al. Mitoxantrone
112. Lowenberg B, Zittoun R, Kerkhofs H, et al. On the value of and cytosine arabinoside as rst-line therapy in elderly pa-
intensive remission induction chemotherapy in elderly patients with acute myeloid leukaemia. Br J Haematol 1991;
tients of 65 » years with acute myeloid leukaemia: a ran-
79: 415 – 20. (P2:104; 86 de novo)

domised phase III study of the European Organisation for
125. Montastruc M, Reiffers J, Stoppa AM, et al. Treatment of
Research and Treatment of Cancer Leukaemia Group. J
acute myeloid leukaemia in elderly patients: the in uence of
Clin Oncol 1989; 7: 1268 – 74. (C2:71)
maintenance therapy (BGM 84 protocol). Nouv Rev Franc
113. Detourmignies L, Wattel E, Lai JL, et al. Is there still a role
d’Hematol 1990; 32: 147 – 52. (P2:92; C3:47)
for low-dose cytosine arabinoside in de novo acute myeloid
126. Rowe JM, Andersen JP, Mazza JJ, et al. A randomised
leukaemia in the elderly? Ann Hematol 1993; 66: 235 – 40.
placebo-controlled phase III study of granulocyte-
(P2:77)
macrophage colonystimulating factor in adult patients ( \55
114. Powell B, Capizzi R, Muss HB, et al. Low-dose ara-C
to 70 years of age) with acute myelogenous leukaemia
therapy for acute myelogenous leukaemia in elderly patients.
(AML): a study of the Eastern Cooperative Oncology Group
Leukaemia 1989; 3: 23 – 8. (P2:44)
115. Tilly H, Castaigne S, Bordessoule D, et al. Low dose cytara- (E1490). Blood 1995; 86: 457 – 62. (C2: 117)
bine vs intensive chemotherapy in the treatment of acute 127. Lowenberg B, Suciu S, Archimbaud E, et al. Use of recom-
non-lymphocytic leukaemia in the elderly. J Clin Oncol 1990; binant GM-CSF during and after remission induction
8: 272 – 9. (C2:126; 87 randomised) chemotherapy in patients aged 61 years and older with acute
116. Baudard M, Marie JP, Cadiou M, et al. Acute myeloid myeloid leukaemia: nal report of AML-11, a phase III
leukaemia in the elderly: retrospective study of 235 consecu- randomised study of the Leukaemia Cooperative Group of
tive patients. Brit J Haematol 1994; 86: 82 – 91. (R1:235) European Organisation for the Research and Treatment of
117. Foon KA, Zighelboim J, Yale C, et al. Intensive chemother- Cancer and the Dutch Belgian Hemato-Oncology Coopera-
apy is the treatment of choice for elderly patients with acute tive Group. Blood 1997; 8: 2952 – 61. (C1:318)
myelogenous leukaemia. Blood 1981; 3: 467 – 70. (R2:107) 128. Stone RM, George SL, Berg DT. Granulocyte macrophage
118. Ruutu T, Almqvist A, Hallman H, et al. Oral induction and colony stimulating factor after initial chemotherapy for el-
consolidation of acute myeloid leukaemia with etoposide, derly patients with primary acute myelogenous leukaemia. N
6-thiogunine, and idarubicin (ETI) in elderly patients: a Engl J Med 1995; 332: 1671 – 7. (C1:388; 347 randomised)
randomised comparison with 5-day TAD. Leukaemia 1994; 129. Heil G, Hoelzer D, Sanz MA, et al. A randomised, double-
8: 11 – 5. (C3:51) blind, placebo-controlled, phase III study of lgrastim in
119. Büchner T, Hiddeman W, Worman B, et al. Daunorubicin remission induction and consolidation therapy for adults
60 instead of 30 mg:m2 improves response and survival in with de novo acute myeloid leukaemia. The International
elderly patients with AML. Blood 1997; 583 (Suppl 1): Abstr Acute Myeloid Leukaemia Study Group. Blood 1997; 90:
504. (C3:340) 4710 – 8. (C1:521)

A Systematic Overview of Chemotherapy Effects in

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Systematic Overview of Chemotherapy Effects in

Uploaded by

Copyright:

Available Formats

SYSTEMATIC OVERVIEW ARTICLE