Treatment of Acute Myelogenous Leukemia
Treatment of Acute Myelogenous Leukemia
There are three broad treatment options for patients
with acute myelogenous leukemia (AML): palliative care only, standard approaches
involving chemotherapy with or without hematopoietic stem cell transplantation,
and investigational approaches studied in the context of a formal clinical
trial, also involving chemotherapy and transplantation. Given the natural
history of the illness (as established by Freireich et al in the 1950s and
1960s), it would be difficult to recommend a purely palliative strategy
except perhaps in the elderly and/or infirm, as discussed below.
The choice between standard and experimental therapy largely depends on the
prognosis following use of the former, and disease prognosis is so variable that
it suggests AML is, in fact, several different diseases. This article considers
the various prognostic subsets and therapies for AML.
In the United States today, most AML patients receive standard therapy under
the care of their private physician as described in this article. The two phases
of standard treatment of AML are induction of a complete remission (CR) and
postremission therapy. Terms such as consolidation, maintenance, or
intensification often are used in connection with postremission therapy.
Consolidation refers to treatment that is only slightly less intense than
induction therapy, maintenance therapy is relatively less intense than
consolidation, and intensification therapy rivals or surpasses induction in dose
The goal of therapy is achievement of a CR (< 5% blasts in the marrow, a
neutrophil count exceeding 1,000-1,500/µL, and a platelet count greater than
100,000/µL); such a remission generally has been assumed necessary, although
not adequate, for prolonged survival. The differences in survival rates between
patients who achieve complete remissions and those who do not have been shown to
be almost entirely dependent upon the time they spend in CR. This observation
suggests that the achievement of CR per se, rather than a difference in inherent
prognosis between responders and nonresponders, is paramount.
The disease recurs in the majority of patients who achieve CRs, making it
likely that not all of these remissions are a result of the same degree of
antileukemic activity. A future clinical challenge will be to distinguish
between significant and cosmetic CR, since the latter predicts especially high
relapse rates unless therapy is changed. Such distinctions eventually will be
made by using polymerase chain reaction (PCR) technology to detect persistent
evidence of a presenting cytogenetic abnormality,[3-5] immunophenotyping to
identify persistent aberrant patterns of cell surface antigens, or simple
clinical observation based on data suggesting that the longer the time required
to achieve CR, the shorter the subsequent CR.[7,8]
At any rate, once 2 years have elapsed from the onset of CR, the risk of
relapse declines precipitously, with this risk continuing to decrease
thereafter. Once 3 years have elapsed from the CR date, the likelihood of
relapse becomes < 10%, and it is reasonable to consider patients potentially
cured at that time.
Standard induction therapy generally consists of a combination of an
anthracycline and cytarabine (ara-C), with the anthracycline often administered
for 3 days and ara-C for 7 days ("3 + 7" regimens). Numerous
randomized studies have compared the efficacy of the anthracyclines daunorubicin
(Cerubidine), mitoxantrone (Novantrone), and idarubicin (Idamycin) when all are
given with 7 days of ara-C.[10-12] In general, most studies, including a large
meta-analysis, found idarubicin to be superior, although differences are
less apparent in older patients.
Several comments about standard anthracycline regimens should be considered:
- Results may vary if different doses (eg, 60 mg/m², rather than 45 or 50
mg/m², of daunorubicin daily) of the several anthracyclines are used.
- Survival gains among the various anthracyclines are relatively modest (ie,
median survivals generally differ by only several months).
- Results from the same treatment regimen (eg, daunorubicin plus ara-C) in
several trials at times exceed the differences among other regimens in the same
These observations make it difficult to attach great significance to the
choice of anthracycline regimens for untreated AML.
Randomized trials have found essentially the same results regardless of the
dose of ara-C (100 or 200 mg/m² daily by continuous infusion for
7 days) or the addition of either thioguanine or etoposide to
the 3 + 7 regimen (with the latter finding obtained in the largest related trial
conducted to date).
Once patients achieve CR, they typically receive several cycles of
consolidation or maintenance therapy employing the initial drugs; treatment is
then stopped and the patient observed. Current data indicate that some
postremission therapy is beneficial, but how much therapy remains unclear, since
durations of therapy are 2 years vs 3 months and 16 vs 7 months in supporting
Again, several comments may be in order. First, benefits generally relate to
disease-free survival rather than to overall survival, and have not been
universally noted.[19,20] Second, the reasons for these discrepancies are
unclear, but they might reflect the myelosuppression produced by the
maintenance, the intensity of the therapy given prior to maintenance, or, given
the variability of AML, the particular mix of patients treated. In any event, we
will assume that prolonging maintenance with these agents for more than 6 months
has relatively little effect on survival, while recognizing that the issue has
not been, and is unlikely to be, unequivocally resolved because of other
Other standard approaches to postremission therapy are allogeneic or
autologous hematopoietic stem cell transplantations, using cyclophosphamide (Cytoxan,
Neosar) plus total-body irradiation or busulfan (Myleran)/cyclophosphamide and high-dose
On average, standard therapy produces median survival durations of 6 to
12 months and CR rates of approximately 60% to 65%; the median duration of
remission is approximately 1 year, with 15% to 20% of the patients potentially
cured. However, these results are inherently misleading, because only a minority
of patients can be considered to conform to an average. For most patients,
outcome is either better or worse than this average, depending on a well-defined
set of prognostic factors.