state of the art review
How I determine if and when to recommend stopping tyrosine
kinase inhibitor treatment for chronic myeloid leukaemia
David M. Ross1,2,3 and Timothy P. Hughes1,2,4
1
Haematology Directorate, SA Pathology, 2School of Medicine, University of Adelaide, Adelaide, 3Flinders University and Medical
Centre, Bedford Park, and 4Cancer Theme, South Australian Health and Medical Research Institute, Adelaide, SA, Australia
Summary
Treatment-free remission (TFR) has recently emerged as a
goal of treatment in chronic myeloid leukaemia (CML).
Molecular remission is sustained in around 30% of imatinib-
treated patients who stop treatment after 2 years with unde-
tectable minimal residual disease (UMRD) by conventional
real-time reverse transcription polymerase chain reaction. An
additional 2030% of patients will lose UMRD, but remain
in stable major molecular remission off treatment. Most
patients with molecular recurrence have a significant increase
in BCR-ABL1 within the first 6 months off treatment, but
there are also rare late relapses. As re-treatment with imati-
nib restores control, a trial of TFR is safe so long as careful
molecular monitoring is provided to enable prompt re-treat-
ment. The minimum eligibility criteria for a trial of TFR are
not yet defined, but the available data support a MRD level
of around a molecular response of 4
5 log for at least
2 years. Factors associated with a higher probability of TFR
include low risk Sokal score, prior interferon treatment,
longer total duration of imatinib treatment and higher num-
bers of natural killer cells at the time of imatinib discontinu-
ation. Preliminary data suggest that the rate of TFR in
patients treated with more potent tyrosine kinase inhibitors
will probably be higher. The biology that underlies TFR is an
area of active investigation.
Keywords: chronic myeloid leukaemia, tyrosine kinase inhib-
itors, minimal residual disease, BCR-ABL1, treatment-free
remission.
What is treatment-free remission?
Some years ago the late Professor John Goldman introduced
the concept of operational cure in chronic phase chronic
myeloid leukaemia (CML) (Goldman & Gordon, 2006). This
Correspondence: Dr David M. Ross, Department of Haematology,
Flinders University and Medical Centre, Bedford Park, SA 5042,
Australia. E-mail: david.ross@health.sa.gov.au
2014 John Wiley & Sons Ltd
British Journal of Haematology, 2014, 166, 311
concept acknowledged the impossibility of proving cure of
CML, whilst also recognizing the enormous clinical success
of a treatment (at the time, imatinib) that enables a patient
to enjoy long-term freedom from symptoms and signs of the
disease, and a normal or near-normal life expectancy, with
minimal toxicity. Treatment-free remission (TFR) takes this
one step further in that patients can enjoy the benefits of dis-
ease control without the need for ongoing treatment.
Although ABL1 tyrosine kinase inhibitors (TKIs) are well-tol-
erated, most patients experience lingering mild toxicity that
can impair quality of life. Furthermore, the newer and more
potent TKIs, such as nilotinib and dasatinib, may be associ-
ated with the emergence of severe adverse events, such as
peripheral vascular events, myocardial ischaemia, pleural
effusions and pulmonary hypertension. Lastly, the cost of
lifelong TKI therapy is substantial. As CML patients now
have a near-normal life expectancy, the provision of TKI
treatment for an average of 25 years is likely to cost around
USD$500 000 per patient, at current prices. If patients are
able to stop treatment after 510 years the potential cost sav-
ing is enormous. For all of these reasons there is considerable
interest among CML patients and their clinicians in the
achievement of TFR.
The first treatment to enable patients to achieve TFR was
allogeneic stem cell transplantation. Patients who achieve
molecular remission post-allograft, without graft-versus-host
disease and without the need for long-term immunosuppres-
sion, have an excellent quality of life and a very low risk of
relapse (Radich et al, 1995; Mughal et al, 2001). Nevertheless,
even this clinical state is not always equivalent to cure,
because very late relapses of CML have been observed (Nor-
kin et al, 2011). The success of allografting, and rescue of
relapse by donor lymphocyte infusion, highlights the poten-
tial importance of immunological surveillance in the long-
term control of CML.
Prior to the advent of imatinib the best non-transplant
therapy for CML was interferon-a (IFN) with or without cyt-
arabine (Guilhot et al, 1997). Only around 20% of IFN-trea-
ted patients achieved a complete cytogenetic response, and a
minority of these patients also achieved undetectable mini-
mal residual disease (UMRD) using sensitive real-time
reverse transcription polymerase chain reaction (RQ-PCR)
First published online 23 April 2014
doi:10.1111/bjh.12892
Review

for BCR-ABL1. Nevertheless, it was shown that TFR (defined
in that study as stable complete cytogenetic response) was
sustained in the majority of these highly-selected patients
after IFN treatment, with a median follow-up of 3 years
(Mahon et al, 2002). The mechanism of action of IFN is
complex and poorly understood, but it is known that IFN
has immunological effects, lending further circumstantial evi-
dence to the idea that the host immune status might be
important in TFR. IFN treatment is associated with the
emergence of cytotoxic T lymphocytes (CTLs) that are reac-
tive against myeloid-associated antigens, and CTLs are asso-
ciated with TFR after IFN treatment (Molldrem et al, 2000;
Burchert et al, 2003; Kanodia et al, 2010). The combination
of IFN and imatinib may have additive or synergistic effects
(Preudhomme et al, 2010).
In 2007 the first report of TFR after imatinib cessation
raised the possibility that prior IFN exposure was associated
with a higher probability of TFR, with neither of the two
patients treated with imatinib de novo (IM-only) remaining
in UMRD after the drug was stopped, versus six out of 10
patients treated initially with IFN, then switched to imatinib
when it became available (IFN-IM) (Rousselot et al, 2007).
The first case of successful TFR after imatinib treatment
alone was reported the following year (Verma et al, 2008)
Subsequently, four prospective clinical trials have reported
TFR in 3050% of patients stopping imatinib after a period
of stable molecular remission (Mahon et al, 2010, 2013; Ross
et al, 2010, 2013; Rousselot et al, 2014). These trials form the
basis of the currently available clinical data; the key features
of the studies are outlined here and summarized in Table I.
Following on from their pilot study in 200405, the
French CML Intergroup conducted the STIM (STop IMati-
nib) study from July 2007 to December 2009 and enrolled
100 patients (Mahon et al, 2010). The key eligibility criteria
were UMRD for a minimum period of 2 years with an esti-
mated RQ-PCR limit of detection (sensitivity) of 5 log.
Around half of the patients received IFN-IM, while the
remainder received IM-only. RQ-PCR monitoring was per-
formed monthly for the first year. Molecular recurrence, the
trigger for resumption of imatinib treatment, was defined as
two consecutive samples with detectable BCR-ABL1 and a
significant rise in BCR-ABL1 level (later defined as 10-fold).
TFR was sustained in 41% of patients at 2 years, with most
molecular recurrence occurring in the first 6 months.
The Australasian Leukaemia and Lymphoma Group
(ALLG) conducted the TWISTER study from August 2006 to
August 2011 and enrolled 40 patients (Ross et al, 2013). As
in STIM, the patients were evenly divided between IM-only
and IFN-IM. Eligibility criteria and monitoring were very
similar to STIM, but the molecular recurrence definition was
more stringent, with any two consecutive positive RQ-PCR
results triggering imatinib re-treatment, even if there was no
rise in BCR-ABL1 from the first to the second test. The esti-
mated rate of TFR at 2 years was 47%.
A French follow-up study, According to STIM (A-STIM)
enrolled an additional 80 similar patients, completing accrual
in June 2012 (Rousselot et al, 2014). A novel element of this
study was that it tested the safety of less stringent molecular
response inclusion criteria (occasional low level BCR-ABL1
was allowed in the preceding 2 years) and a less stringent
molecular recurrence definition, loss of major molecular
response (MMR), i.e. BCR-ABL1 >0
1%. This enabled the
TFR rate to be described for a similar population using
recurrence definitions of varying stringency. Using the STIM
definition, the rate of TFR at 2 years was 46%, while using
the new definition (loss of MMR) the rate of TFR at 2 years
was 64%.
The most recent study from the French CML Intergroup,
STIM2, has not yet been published, but was recently pre-
sented at the American Society of Hematology annual meet-
ing (Mahon et al, 2013). This study is important because in
the current era the overwhelming majority of CML patients
are treated only with TKIs. STIM2 enrolled 124 IM-only
patients between April 2011 and June 2013. Using the origi-
nal STIM definition of molecular recurrence, 61% of patients
were in TFR at the time of the interim analysis (median fol-
low-up 12 months). Importantly, 33% of patients in TFR
showed fluctuating low levels of BCR-ABL1 below 0
1%. It is
not clear why the proportion of patients in STIM2 with loss
of UMRD, but stable MMR is higher than in the previous
studies, but taking this into account, the rate of TFR defined
by the stringent UMRD criterion is similar to that seen in

Table I. Summary of prospective trials of imatinib discontinuation.

Rate of TFR according to different recurrence definitions*

% IM- Median Median duration Median duration Loss of UMRD 10-fold Loss of
Study N only age (years) of imatinib (months) of UMRD (months) (TWISTER) rise (STIM) MMR (A-STIM)

STIM 100 51 62 50 36 (2485) NR 41 NR

TWISTER 40 48 61 70 36 (2482) 47 NR NR
A-STIM 80 46 55 79 41 (2496) NR 44 61
STIM2 124 100 61 NR NR 34 61 NR

IM, imatinib; UMRD, undetectable minimal residual disease; TFR, treatment-free remission; MMR, major molecular response; NR, not reported.
*Rates of TFR are not directly comparable between studies as estimates may refer to different time points. STIM2 is in progress and only interim
results are available.
As intermittent low positive results were allowed this is not strictly a duration of UMRD.

4 2014 John Wiley & Sons Ltd

British Journal of Haematology, 2014, 166, 311

IM-only patients in TWISTER, and in a small retrospective
series from Korea, at around 30% (Yhim et al, 2012).
Which patients are eligible for a trial of TFR?
We estimate that around 40% of patients treated with imati-
nib de novo will eventually be eligible for a trial of TKI dis-
continuation, based on the criteria used in STIM and
TWISTER (Branford et al, 2013). Therefore, approximately
1015% of all patients starting imatinib for CML can expect,
at some time in the future, to be able to remain off therapy
with UMRD. These percentages may change over time as the
eligibility criteria for TKI discontinuation and the criteria for
resumption of TKI therapy are refined.
When the first TKI discontinuation studies were planned
there were no data on the safety of this approach. In order
to select the patients with the most favourable prognosis the
first studies of TFR included only patients with UMRD
[approximately equivalent to a molecular response of 4
5 log
(MR 4
5) or BCR-ABL1 0
003%] sustained for at least
2 years. In this population (a mixture of IM-only and IFN-
IM) the rate of stable UMRD off treatment was approxi-
mately 40%.
The apparently dichotomous outcome of either early
recurrence or durable TFR raised the question of whether or
not the leukaemic clone was eradicated in some patients.
Given that the level of BCR-ABL1 may continue to fall on
TKI treatment after achieving UMRD, the actual level of
MRD in these patients could be anywhere from just below
0
003% to zero. Is the difference between the 40% in TFR
and the 60% with molecular recurrence simply a product of
the depth of MRD at the time of stopping? Two lines of evi-
dence suggest that it is not. Kinetic studies have shown a
slow, progressive depletion of MRD during continued imati-
nib treatment (Michor et al, 2005; Roeder et al, 2006), so
one would expect that a longer duration of imatinib after the
first achievement of UMRD would be associated with a lower
risk of molecular recurrence, but that has not proven to be
the case. Secondly, highly sensitive BCR-ABL1 DNA PCR
with a limit of detection of around 1-log below that of con-
ventional RQ-PCR detected MRD in the majority of patients
in the TWISTER study, even in those patients who have
remained in TFR for several years (Ross et al, 2010, 2013).
Using RQ-PCR on sorted cell populations, Rousselot et al
(2014) showed that the residual BCR-ABL1 signal in deep
molecular response arises from the granulocyte lineage.
Therefore, the persistence of BCR-ABL1 positivity in patients
with TFR cannot be explained simply by persistence of long-
lived lymphocytes, which was one hypothesis raised by these
original findings.
UMRD is essentially an arbitrary line in the sand. The
commonly used sensitivity requirement of 4
5-log represents
an MRD threshold that is based on the technical limitations
of RQ-PCR technology, not based on empirical data.
This raises the question of what level of MRD really is a
2014 John Wiley & Sons Ltd
British Journal of Haematology, 2014, 166, 311
Review
prerequisite for TFR. A-STIM enrolled patients who had an
occasional low level positive RQ-PCR test during the qualify-
ing period, and this slight relaxation of the MRD inclusion
criterion did not have a measurable effect on the TFR rate.
Several studies currently underway have used MR4.5 (which
includes both UMRD and patients with low level detectable
BCR-ABL1) as an inclusion criterion that is more easily
standardized, and the results of this next generation of TFR
studies will soon be available.
The studies outlined above have generally included only
patients in chronic phase, with no history of BCR-ABL1
kinase domain mutations. STIM allowed the enrolment of
patients with accelerated phase CML, although the number
of such patients was not reported. A series of three patients
who stopped dasatinib in UMRD after a history of imatinib
intolerance or failure (including BCR-ABL1 kinase domain
mutations) was reported (Ross et al, 2011). Two of the three
patients remained in UMRD off treatment, while the third
patient resumed dasatinib after molecular recurrence, and
regained UMRD. Until further data are available we would
not recommend stopping TKIs in patients with a history of
advanced phase disease or kinase domain mutations, unless
in the context of a clinical trial.
Patient age may be a factor in deciding whether or not to
stop TKI. Older patients may be less concerned about the
potential for late toxicity from treatment, including effects
on fertility or teratogenicity. Younger patients may experi-
ence a greater loss of quality of life from TKI toxicity (Effic-
ace et al, 2011) and may have a greater immediate incentive
to stop treatment. The age range of patients in TFR studies
reported to date is 2784 years, and age has not emerged as
a factor influencing molecular recurrence (Mahon et al,
2010, 2013; Ross et al, 2013; Rousselot et al, 2014). It should
be noted, however, that advanced age increases the Sokal risk
score (Sokal et al, 1984), which may influence recurrence risk
(see below).
How should patients be monitored off TKI
therapy?
The availability of a sensitive RQ-PCR assay with a rapid
turn-around time is essential for patients to safely undergo a
trial of TFR. Monitoring with karyotyping or fluorescence in
situ hybridization is not sufficient, because most clinicians
would re-start TKI at loss of MMR. If an insensitive assay is
used the UMRD threshold may actually be little better than
MMR, and the risk of molecular recurrence and adverse out-
comes may well be higher in patients who stop with a higher
disease burden. Considerable effort has been expended
around the world to standardize reporting of RQ-PCR detec-
tion limits, and to make available good quality and highly
sensitive RQ-PCR tests.
Most patients who experience molecular recurrence after
TKI withdrawal do so within the first 6 months (Mahon
et al, 2010; Ross et al, 2013). The median time to detectable
5
Review
BCR-ABL1 is 34 months, and almost all patients relapsing
early do so with exponentially rising BCR-ABL1 levels, so
that without additional treatment loss of MMR would occur
within 12 months of losing UMRD (Branford et al, 2012).
Therefore, we recommend monitoring monthly for
12 months with a sensitive RQ-PCR assay. Less frequent
monitoring (e.g. every 23 months) may be sufficient there-
after, but any positive result should trigger repeat RQ-PCR
the following month to confirm the abnormality and to
assess the kinetics of BCR-ABL1. The latest molecular recur-
rence reported to date was at 42 months in a Japanese retro-
spective survey of imatinib discontinuation (Takahashi et al,
2012). The latest relapse in a prospective clinical trial was in
the TWISTER study at 27 months: that patient had two con-
secutive BCR-ABL1 values of 0
002% and the following
month again had UMRD before imatinib was re-started
(Ross et al, 2013). In A-STIM, the latest loss of MMR was at
17 months (Rousselot et al, 2014).
Molecular recurrence of CML after TKI withdrawal has
not been associated with kinase domain mutations. In the
TWISTER study, kinase domain mutation analysis was per-
formed in all patients with detectable BCR-ABL1 and no
mutations were detected (Ross et al, 2013). This is perhaps
not surprising, given that BCR-ABL1 mutant clones do not
have a proliferative advantage over wild-type BCR-ABL1
unless the relevant TKI is present.
What is failure of TFR, and how should it be
managed?
The concept of failure, like the concept of an optimal
response, is moveable. Failure of TFR after IFN treatment
was loss of complete cytogenetic response, in A-STIM it was
loss of MMR, and in TWISTER it was loss of UMRD.
Whether or not the less stringent recurrence definitions will
result in TFR in the longer term remains to be seen. In the
shorter term it will be interesting to hear patients perspec-
tives on whether this represents a therapeutic success. After
spending years striving for a deep molecular response, and
eventual UMRD, our patients may be somewhat confused by
our acceptance of MMR without therapy. Patient education
and counselling are an important part of the clinicians role
in TFR.
The average doubling time of BCR-ABL1 for patients off
TKI therapy is around 10 d, equating to an increase of
around 1-log every month (Branford et al, 2012). This dou-
bling time is observed in most patients who fail a trial of
TFR. In patients with exponentially rising BCR-ABL1 one
can expect loss of MMR to occur approximately 2 months
after BCR-ABL1 is first detected. We would recommend
re-starting TKI in any patient who shows a short doubling
time, rather than waiting for loss of MMR.
All of the TKI discontinuation studies to date have used
molecular recurrence as a trigger to re-start treatment, and
in all cases patients have remained responsive to treatment.
6 2014 John Wiley & Sons Ltd
The overwhelming majority of patients regain UMRD, typi-
cally within 6 months. In a single case a patient who had
molecular recurrence after ceasing imatinib and subsequently
regained MMR on imatinib abruptly developed blast crisis
9 months later. That patient was in the IFN-IM cohort of
A-STIM and started imatinib 7 years after her original diag-
nosis of CML (Rousselot et al, 2014).
A small number of patients in STIM and TWISTER
stopped imatinib for a second time after regaining UMRD,
and remaining on treatment in UMRD for at least
12 months (Legros et al, 2012; Ross et al, 2013). Interest-
ingly, a second molecular recurrence was not inevitable, and
around 25% of patients remained off treatment without loss
of MMR.
What factors predict durable TFR?
Several themes are emerging from the data available so far
when trying to predict TFR. Receiving more than 50 months
of imatinib therapy before the trial of TFR was one of the
strongest predictors of relapse risk (hazard ratio 0
421)
(Mahon et al, 2010). The Sokal risk score calculated at diag-
nosis has been consistently identified as prognostic (Mahon
et al, 2010; Yhim et al, 2012; Ross et al, 2013). Among Sokal
low risk patients, 5060% sustained TFR, whereas in high
risk patients the rate of TFR was around 1020%, suggesting
that the effect of adverse disease biology at diagnosis is not
completely overcome in TKI-responsive patients, even in this
highly selected fraction of patients with the best possible
response to treatment.
The prognostic value of the initial and subsequent fall in
the level of BCR-ABL1 in response to imatinib treatment has
not been examined in great detail. STIM, A-STIM and
TWISTER all enrolled a mixture of patients treated with IM-
only and IFN-IM, resulting in heterogeneous response kinet-
ics. In the relatively small number of IM-only patients whose
BCR-ABL1 kinetics could be examined there was no link
between the time taken to achieve molecular response targets
and the later probability of TFR (Mahon et al, 2010; Ross
et al, 2013; Rousselot et al, 2014).
A trend seen in all three studies that included IFN-IM
patients was a higher rate of TFR in that cohort than in IM-
only patients: around 50% vs. around 30%. There was a
trend toward higher TFR rate in patients with a longer dura-
tion of IFN exposure (Ross et al, 2013). A beneficial effect of
prior IFN treatment was also observed in a retrospective
study from Japan (Takahashi et al, 2012). There are various
hypotheses that might account for a beneficial effect of IFN,
including its pleiotropic immunological effects and stem cell
depletion (Essers et al, 2009). However, there is an inherent
selection bias in the IFN-IM cohort as, historically, only a
minority of patients remained on IFN for long enough to
benefit from switching to IM when it first became available,
and biological low risk patients are likely to be over-repre-
sented in this cohort. Burchert et al (2010) reported a
British Journal of Haematology, 2014, 166, 311

number of patients treated with the combination of imatinib
and pegylated IFN, following which imatinib was stopped
and IFN continued as maintenance treatment. Whilst this
is a slightly different clinical setting it demonstrated a pro-
gressive reduction in MRD level after the withdrawal of
TKI, associated with the emergence of CTLs against CML-
associated antigens.
We started the TWISTER study with the hypothesis that
the depth of MRD at the time of stopping imatinib would be
the major determinant of relapse risk. Consequently, we
developed a highly sensitive (6-log below baseline) patient-
specific assay for BCR-ABL1 DNA, with the aim of stratifying
patients according to depth of response and risk of relapse
(Ross et al, 2010). Unexpectedly, we found that the majority
of patients had detectable BCR-ABL1 DNA at the time of
stopping TKI. Consequently, the presence or absence of
MRD was not informative with regard to subsequent molec-
ular recurrence, although it did enable the detection of
molecular recurrence 12 months earlier than RQ-PCR.
Technological advances in PCR continuously re-define
what can be achieved in MRD analysis. The difficulty of
detecting BCR-ABL1 breakpoints is no longer a major obsta-
cle with several methods now available (based on multiplex
PCR or massively parallel paired-ends sequencing) (Bartley
et al, 2010; Score et al, 2010; Alikian et al, 2013). For a long
time there has been speculation that CML cells in MRD may
be quiescent and have reduced expression of BCR-ABL1, and
this could give a potential advantage to DNA PCR (Kumari
et al, 2012). However, there are conflicting data on this
topic: BCR-ABL1 mRNA is enriched in the granulocyte com-
partment of CML patients with UMRD (Rousselot et al,
2014), and primitive stem cells isolated in vitro still express
BCR-ABL1 transcripts (Copland et al, 2006). In our opinion,
the major advantage of BCR-ABL1 DNA is its specificity, so
1. Induction of 2. Consolidation
molecular of molecular
remission remission
More potent TKIs Increased
duration of TKI
Combination of
prior to stopping
TKI with other
drugs (e.g. IFN) Development of
targetting immune biomarkers for
responses or CML recurrence risk
stem cells
Fig 1. Possible strategies to maximize the number of patients in TFR. CML, chronic myeloid leukaemia; TFR, treatment-free remission;
TKI, tyrosine kinase inhibitor; IFN, interferon-a.
2014 John Wiley & Sons Ltd
British Journal of Haematology, 2014, 166, 311
Review
that DNA PCR is much less susceptible to false positive
results. There is no doubt that increased sensitivity can be
achieved by replicate PCR testing (for both RNA and DNA
targets), and an important development in this regard is dig-
ital PCR (dPCR). In dPCR the reaction mixture is parti-
tioned into multiple tiny volumes (either chambers in a
microfluidic system or droplets in an emulsion). In an MRD
setting, where the number of targets is small, each dPCR
reaction will contain a maximum of one copy of the target,
so every result is positive or negative (hence, digital). Stud-
ies in CML or Ph-positive ALL have shown the feasibility of
this approach for detecting low levels of BCR-ABL1, and at
least one study is examining its utility in the TFR setting
(Goh et al, 2011; Mori et al, 2013; Iacobucci et al, 2014).
What are the key biological questions in the
field of TFR?
We consider TFR to be a distinct biological state that can be
achieved in a subset of patients with deep molecular
response. The pathobiology of TFR remains to be defined,
but the two most promising fields of investigation are the
properties of the CML clone, and the nature of the host
immune response. Figure 1 summarizes strategies that could
be used to maximize the numbers of CML patients who
eventually achieve TFR.
The main evidence that we have at present in relation to
how the properties of the CML clone influence the likelihood
of TFR is the association with Sokal score. Sokal, and other
clinical risk scores, must have a biological basis. Gene expres-
sion studies have shown alterations in a number of genes to be
associated with Sokal risk score (Schmidt et al, 2001; Flamant
et al, 2010), but the biological basis of high-risk disease has
not yet been characterized in detail. The study of CML
3. New strategies
Failure of for second attempt
TFR at TFR
More potent TKIs
Combination of TKI
with drugs targetting
immune responses or
CML stem cells
Successful 3. Long-term
TFR molecular
monitoring
7
Review
disease-specific properties in TFR patients is hampered by two
practical problems. Firstly, the median time from diagnosis to
TKI discontinuation is usually >5 years, so cryopreserved cells
from diagnosis are typically not available for study; and sec-
ondly, the number of persistent leukaemic cells at the time of
stopping TKI treatment is too small to permit isolation of suf-
ficient numbers for study. Clinical trials in which patients are
treated de novo, and then have the option of proceeding to
TFR will represent the best opportunity to provide detailed
information on how the properties of the original CML clone
influence TFR outcome.
At present the only reported potential biomarkers for suc-
cessful TFR are related to immunological status. Immunolog-
ical anergy is a feature of CML, affecting both T cells and
natural killer (NK) cell activity (Chen et al, 2008, 2012;
Mumprecht et al, 2009). One study found increased numbers
of IFN-c-producing NK cells and decreased numbers of nave
T cells (CD8+ CD62L+) in patients with TFR after imatinib
when compared with patients with UMRD remaining on
imatinib treatment (Ohyashiki et al, 2012). As patients in
TFR were not studied before imatinib was withdrawn it is
unknown whether this phenotype has any prognostic value,
but follow-up studies are warranted. Rea et al (2013) pre-
sented the results of immunological studies in a subset of 51
patients enrolled in the STIM study. Average NK cell num-
bers in the peripheral blood at the time of stopping imatinib
were higher in patients who sustained TFR than in those
with molecular recurrence. The difference was primarily due
to the differentiated CD56-dim compartment that has the
greatest cytotoxic activity, and accounts for most of the NK
cells in the blood of normal individuals. Interestingly, there
was no increase in NK cell numbers when patients were
taken off imatinib, arguing that this immunological defect is
not simply a toxicity of the TKI, but is related in some way
to the disease or to constitutional factors. There was consid-
erable overlap between NK cell numbers in TFR patients and
in patients destined for recurrence. More detailed immuno-
logical profiling might help to improve risk stratification.
What are the key unanswered clinical
questions in the field of TFR?
Many important clinical questions remain regarding the selec-
tion of patients for a trial of TFR. What is the optimum level of
MRD at which a trial of TFR should be entertained? Is there a
continuous gradient with higher TFR rate as the level of MRD
falls, or is there a critical threshold below which there is no fur-
ther improvement in the probability of TFR? In A-STIM the
inclusion criteria were relaxed to allow patients with intermit-
tent low-level positive RQ-PCR tests during the 2 years prior
to stopping imatinib. The rate of loss of UMRD (STIM defini-
tion) was higher in these patients, but there was no difference
in the rate of loss of MMR (Rousselot et al, 2014). A single
centre study reported on the outcome of five patients who
stopped TKI with MR4.5 but detectable MRD: four patients
8 2014 John Wiley & Sons Ltd
lost MR4.5, but only one patient lost MMR (Benjamini et al,
2013).
What is the minimum duration of MR4.5 (or UMRD) before
attempting TFR? The retrospective series from Japan reported
on the outcome of 43 patients who stopped imatinib with
UMRD and remained off treatment for at least 6 months
(Takahashi et al, 2012). The median duration of UMRD prior
to imatinib cessation in patients who sustained TFR was
32
5 months vs. 6 months in patients with a molecular recur-
rence. The estimated rate of TFR at 2 years was almost 80% in
those with at least 2 years of UMRD prior to TKI discontinua-
tion versus 30% if the period of UMRD was shorter. These
figures are not directly comparable with those from the pro-
spective studies, because the requirement for patients to
remain off TKI for at least 6 months introduces a bias against
patients with early molecular recurrence. A report from the
MD Anderson Cancer Center found that patients with UMRD
for more than 5 years at the time of TKI discontinuation had a
higher rate of TFR (Benjamini et al, 2013).
What duration of BCR-ABL1 monitoring is required in
CML patients in TFR? It should be emphasized that the
number of patients with long-term follow-up remains very
small, and there is still (at least theoretically) a risk of molec-
ular recurrence even after five or more years. Although most
molecular recurrence occurs in the first 6 months, the num-
ber of instances of loss of UMRD in the second and third
years is not trivial. The frequency of molecular monitoring
may be reduced after the second year of TFR, but we would
recommend indefinite RQ-PCR testing at least every
3 months. This is important both to establish the long-term
durability of TFR, and to safeguard each individual patient
against the delayed diagnosis of relapse. In our practice the
cost of an RQ-PCR test is equivalent to around 1 week of
imatinib treatment, so RQ-PCR monitoring of TFR remains
highly cost-effective.
Perhaps the most pressing question is whether there is a
difference in the rate of TFR after dasatinib or nilotinib treat-
ment. Early response data from the pivotal up-front studies
indicate that these more potent TKIs will increase the pro-
portion of patients who meet the eligibility criteria for a trial
of TFR (Kantarjian et al, 2012; Larson et al, 2012). The
ENESTcmr study also showed that switching patients from
imatinib to nilotinib could increase the rate of MR4.5 (Leber
et al, 2013). However, the achievement of deep molecular
response is necessary for TFR, but not sufficient, and it
remains to be seen what rate of TFR is achieved in those
patients. Stopping studies in patients treated with nilotinib
first-line or after imatinib (ENESTfreedom, ENESTop) and
dasatinib (DASFREE) are underway, but the results from
these large international studies are not yet available. A
French study of stopping second-generation (2G) TKIs
reported on the outcome of 34 patients with at least
6 months of follow-up (Rea et al, 2012). TFR (defined as
persistent MMR off treatment) was observed in 58% of
patients, or 44% using the stringent TWISTER criterion.
British Journal of Haematology, 2014, 166, 311
 Review

Around half of the patients switched from imatinib to a 2G
TKI due to intolerance, and around half for suboptimal
response. The interim presentation of a Japanese dasatinib
discontinuation study reported TFR (stable UMRD) in 44%
of 63 patients with UMRD after second-line dasatinib
(Tanaka et al, 2013). Based on these early data it seems that
the rate of TFR after 2G TKIs may be similar to the rate seen
after imatinib or slightly better, but if more patients achieve
a deep molecular response, the absolute number of patients
who achieve TFR could be substantially higher than in
IM-only patients.
Is there a novel treatment that could be added to a TKI to
increase the probability of TFR in patients who already have a
deep molecular response, either before stopping TKI or as a
second-line treatment after an initial failure of TKI? There
are several candidate drugs in this area: e.g. immunomodula-
tion with IFN or ipilimumab (Bashey et al, 2009), stem cell
depletion with IFN or inhibitors of the Hedgehog pathway
(Zhang et al, 2010), or targeting of alternative signalling
pathways, such as Janus kinase (JAK) or phosphatidylinosi-
tide 3-kinase (PI3K) (Ding et al, 2013; Warsch et al, 2013).
Given that patients with a deep molecular response on TKIs
generally have good quality of life and a normal life expec-
tancy the potential toxicity of any new therapy must be care-
fully considered.
Stopping TKIs outside the setting of a clinical
trial
Wherever possible, patients wishing to discontinue TKI treat-
ment should do so in the structured setting of a clinical trial.
This provides a degree of protection to the patient and adds
to the body of scientific knowledge available. In cases where
no such opportunity exists, a patient may still wish to dis-
continue treatment. We consider that this can safely be
undertaken under the supervision of an experienced clini-
cian, so long as high quality RQ-PCR test results are avail-
able within a reasonable length of time. The patient must
understand the need for more intensive follow-up with
molecular testing, and be prepared to comply with the fol-
low-up plan. Inadequate follow-up testing could expose the
patient to unnecessary risk.
For an individual patient the decision whether or not to dis-
continue TKI is dependent on the motivation to stop treat-
ment and the estimated risk of molecular recurrence. What is
known about prediction of molecular recurrence off therapy
has been summarized. Among patients with stable MR4.5 or
UMRD for at least 2 years there is no identified subgroup of
patients with a TFR rate below 10%. Factors that may drive an
attempt at TKI discontinuation include: (i) patient preference;
(ii) significant TKI toxicity; (iii) cost or (iv) planned preg-
nancy. In a patient who has a high likelihood of molecular
recurrence (e.g. high risk Sokal) a trial of TFR may not be jus-
tified unless there is a strong motivation for the patient to dis-
continue therapy. Although the clinical risks associated with
short-term molecular recurrence are minor, an unsuccessful
attempt at TKI discontinuation uses additional monitoring
resources, and may cause unnecessary anxiety.
Concluding remarks
As TFR has become a possibility for many CML patients the
goal of treatment is expanding beyond simply preventing dis-
ease progression to achieving the deepest possible molecular
response in order to enable the patient to be free of treat-
ment and free of disease. This trend has parallels in other
haematological neoplasms (e.g. limited stage Hodgkin lym-
phoma or paediatric acute lymphoblastic leukaemia) that
respond so well to treatment that recent clinical trials have
focussed on the de-escalation of therapy to try to minimize
the late toxicities of treatment without compromising effi-
cacy. With this new aim come new risks: the single case of
blast crisis after a trial of TFR reminds us that careful moni-
toring is essential. At least for now, most attempts at TKI
discontinuation should, if possible, be undertaken in the set-
ting of a clinical trial. Just as it seemed that all the major
questions in CML treatment were close to being answered a
whole new area of investigation has emerged.
Author contributions
DMR and TPH reviewed the literature and wrote the paper.
Disclosures
DMR and TPH were the principal investigators in the TWIS-
TER study. DMR has received research funding and honor-
aria from Novartis Pharmaceuticals. TPH has received
research funding and honoraria from Novartis, Bristol-Myers
Squibb, and Ariad.

References
Alikian, M., Ellery, P., Forbes, M., Gerrard, G.,
Aitman, T., Kasperaviciute, D., Milojkovic, D.,
Marin, D., Goldman, J.M., Apperley, J.F.,
Foroni, L. & Reid, A. (2013) NGS-assisted
DNA-based digital qPCR facilitates stratification
of CML patients in long-term molecular remis-
sion based on the presence of detectable BCR-
ABL1 DNA. Blood, 122, Abstract 4006.
Bartley, P.A., Martin-Harris, M.H., Budgen, B.J.,
Ross, D.M. & Morley, A.A. (2010) Rapid isola-
tion of translocation breakpoints in chronic
myeloid and acute promyelocytic leukaemia.
British Journal of Haematology, 149, 231236.
Bashey, A., Medina, B., Corringham, S., Pasek, M.,
Carrier, E., Vrooman, L., Lowy, I., Solomon,
S.R., Morris, L.E., Holland, H.K., Mason, J.R.,
Alyea, E.P., Soiffer, R.J. & Ball, E.D. (2009)
CTLA4 blockade with ipilimumab to treat
relapse of malignancy after allogeneic hemato-
poietic cell transplantation. Blood, 113, 1581
1588.
Benjamini, O., Kantarjian, H., Rios, M.B., Jabbour,
E., OBrien, S., Jain, P., Cardenas-Turanzas, M.,
Faderl, S., Garcia-Manero, G., Ravandi, F.,
Borthakur, G., Quintas-Cardama, A. & Cortes, J.
(2013) Patient-driven discontinuation of

2014 John Wiley & Sons Ltd 9

British Journal of Haematology, 2014, 166, 311
Review
tyrosine kinase inhibitors: single institution
experience. Leukaemia & Lymphoma, doi: 10.
3109/10428194.2013.831092.
Branford, S., Yeung, D.T., Prime, J.A., Choi, S.Y.,
Bang, J.H., Park, J.E., Kim, D.W., Ross, D.M. &
Hughes, T.P. (2012) BCR-ABL1 doubling times
more reliably assess the dynamics of CML
relapse compared with the BCR-ABL1 fold rise:
implications for monitoring and management.
Blood, 119, 42644271.
Branford, S., Yeung, D.T., Ross, D.M., Prime, J.A.,
Field, C.R., Altamura, H.K., Yeoman, A.L.,
Georgievski, J., Jamison, B.A., Phillis, S., Sulli-
van, B., Briggs, N.E., Hertzberg, M., Seymour,
J.F., Reynolds, J. & Hughes, T.P. (2013) Early
molecular response and female sex strongly pre-
dict stable undetectable BCR-ABL1, the criteria
for imatinib discontinuation in patients with
CML. Blood, 121, 38183824.
Burchert, A., Wolfl, S., Schmidt, M., Brendel, C.,
Denecke, B., Cai, D., Odyvanova, L., Lahaye, T.,
Muller, M.C., Berg, T., Gschaidmeier, H., Wit-
tig, B., Hehlmann, R., Hochhaus, A. & Neu-
bauer, A. (2003) Interferon-alpha, but not the
ABL-kinase inhibitor imatinib (STI571), induces
expression of myeloblastin and a specific T-cell
response in chronic myeloid leukemia. Blood,
101, 259264.
Burchert, A., Muller, M.C., Kostrewa, P., Erben,
P., Bostel, T., Liebler, S., Hehlmann, R., Neu-
bauer, A. & Hochhaus, A. (2010) Sustained
molecular response with interferon alfa mainte-
nance after induction therapy with imatinib plus
interferon alfa in patients with chronic myeloid
leukemia. Journal of Clinical Oncology, 28, 1429
1435.
Chen, C.I., Maecker, H.T. & Lee, P.P. (2008)
Development and dynamics of robust T-cell
responses to CML under imatinib treatment.
Blood, 111, 53425349.
Chen, C.I., Koschmieder, S., Kerstiens, L., Schemi-
onek, M., Altvater, B., Pscherer, S., Gerss, J.,
Maecker, H.T., Berdel, W.E., Juergens, H., Lee,
P.P. & Rossig, C. (2012) NK cells are dysfunc-
tional in human chronic myelogenous leukemia
before and on imatinib treatment and in BCR-
ABL-positive mice. Leukemia, 26, 465474.
Copland, M., Hamilton, A., Elrick, L.J., Baird,
J.W., Allan, E.K., Jordanides, N., Barow, M.,
Mountford, J.C. & Holyoake, T.L. (2006) Dasat-
inib (BMS-354825) targets an earlier progenitor
population than imatinib in primary CML but
does not eliminate the quiescent fraction. Blood,
107, 45324539.
Ding, J., Romani, J., Zaborski, M., Macleod, R.A.,
Nagel, S., Drexler, H.G. & Quentmeier, H.
(2013) Inhibition of PI3K/mTOR overcomes nil-
otinib resistance in BCR-ABL1 positive leukemia
cells through translational down-regulation of
MDM2. PLoS ONE, 8, e83510.
Efficace, F., Baccarani, M., Breccia, M., Alimena,
G., Rosti, G., Cottone, F., Deliliers, G.L., Barate,
C., Rossi, A.R., Fioritoni, G., Luciano, L., Turri,
D., Martino, B., Di Raimondo, F., Dabusti, M.,
Bergamaschi, M., Leoni, P., Simula, M.P., Leva-
10
to, L., Ulisciani, S., Veneri, D., Sica, S., Rambal-
di, A., Vignetti, M., Mandelli, F. & GIMEMA
(2011) Health-related quality of life in chronic
myeloid leukemia patients receiving long-term
therapy with imatinib compared with the gen-
eral population. Blood, 118, 45544560.
Essers, M.A., Offner, S., Blanco-Bose, W.E., Wai-
bler, Z., Kalinke, U., Duchosal, M.A. & Trumpp,
A. (2009) IFNalpha activates dormant haemato-
poietic stem cells in vivo. Nature, 458, 904908.
Flamant, S., Ritchie, W., Guilhot, J., Holst, J., Bon-
net, M.L., Chomel, J.C., Guilhot, F., Turhan,
A.G. & Rasko, J.E. (2010) Micro-RNA response
to imatinib mesylate in patients with chronic
myeloid leukemia. Haematologica, 95, 1325
1333.
Goh, H.G., Lin, M., Fukushima, T., Saglio, G.,
Kim, D., Choi, S.Y., Kim, S.H., Lee, J., Lee, Y.S.,
Oh, S.M. & Kim, D.W. (2011) Sensitive quanti-
tation of minimal residual disease in chronic
myeloid leukemia using nanofluidic digital poly-
merase chain reaction assay. Leukaemia & Lym-
phoma, 52, 896904.
Goldman, J. & Gordon, M. (2006) Why do
chronic myelogenous leukemia stem cells survive
allogeneic stem cell transplantation or imatinib:
does it really matter? Leukaemia & Lymphoma,
47, 17.
Guilhot, F., Chastang, C., Michallet, M., Guerci,
A., Harousseau, J.L., Maloisel, F., Bouabdallah,
R., Guyotat, D., Cheron, N., Nicolini, F., Abgr-
all, J.F. & Tanzer, J. (1997) Interferon alfa-2b
combined with cytarabine versus interferon
alone in chronic myelogenous leukemia. French
Chronic Myeloid Leukemia Study Group. New
England Journal of Medicine, 337, 223229.
Iacobucci, I., Lonetti, A., Venturi, C., Ferrari, A.,
Papayannidis, C., Ottaviani, E., Abbenante,
M.C., Paolini, S., Bresciani, P., Potenza, L.,
Parisi, S., Cattina, F., Soverini, S., Russo, D.,
Luppi, M. & Martinelli, G. (2014) Use of a high
sensitive nanofluidic array for the detection
of rare copies of BCR-ABL1 transcript in
patients with Philadelphia-positive acute lym-
phoblastic leukemia in complete response. Leu-
kemia Research, doi: 10.1016/j.leukres.2014.02.
005.
Kanodia, S., Wieder, E., Lu, S., Talpaz, M., Ala-
trash, G., Clise-Dwyer, K. & Molldrem, J.J.
(2010) PR1-specific T cells are associated with
unmaintained cytogenetic remission of chronic
myelogenous leukemia after interferon with-
drawal. PLoS ONE, 5, e11770.
Kantarjian, H.M., Shah, N.P., Cortes, J.E., Baccara-
ni, M., Agarwal, M.B., Undurraga, M.S., Wang,
J., Ipina, J.J., Kim, D.W., Ogura, M., Pavlovsky,
C., Junghanss, C., Milone, J.H., Nicolini, F.E.,
Robak, T., Van Droogenbroeck, J., Vellenga, E.,
Bradley-Garelik, M.B., Zhu, C. & Hochhaus, A.
(2012) Dasatinib or imatinib in newly diagnosed
chronic-phase chronic myeloid leukemia: 2-year
follow-up from a randomized phase 3 trial
(DASISION). Blood, 119, 11231129.
Kumari, A., Brendel, C., Hochhaus, A., Neubauer,
A. & Burchert, A. (2012) Low BCR-ABL expres-
British Journal of Haematology, 2014, 166, 311
sion levels in hematopoietic precursor cells
enable persistence of chronic myeloid leukemia
under imatinib. Blood, 119, 530539.
Larson, R.A., Hochhaus, A., Hughes, T.P., Clark,
R.E., Etienne, G., Kim, D.W., Flinn, I.W.,
Kurokawa, M., Moiraghi, B., Yu, R., Blakesley,
R.E., Gallagher, N.J., Saglio, G. & Kantarjian,
H.M. (2012) Nilotinib vs imatinib in patients
with newly diagnosed Philadelphia chromo-
some-positive chronic myeloid leukemia in
chronic phase: ENESTnd 3-year follow-up. Leu-
kemia, 26, 21972203.
Leber, B., Cervantes, F., Spector, N., Lipton, J.H.,
Etienne, G., Pasquini, R., Clementino, N.C.,
Guerci-Bresler, A., Forrest, D., Schwarer, A.P.,
Acharya, S., Collins, L., Szczudlo, T. & Hughes,
T.P. (2013) Achievement and maintenance of
deeper molecular response by switching to nil-
otinib in patients (pts) with chronic myeloid
leukemia in chronic phase (CML-CP) with
residual disease on long-term imatinib: ENES-
Tcmr 36-month follow-up. Blood (ASH Annual
Meeting Abstracts), 120, 94.
Legros, L., Rousselot, P., Giraudier, S., Tulliez, M.,
Huguet, F., Nicolini, F.E. & Mahon, F.X. (2012)
Second attempt to discontinue imatinib in CP-
CML patients with a second sustained complete
molecular response. Blood, 120, 19591960.
Mahon, F.X., Delbrel, X., Cony-Makhoul, P., Fab-
eres, C., Boiron, J.M., Barthe, C., Bilhou-Nabera,
C., Pigneux, A., Marit, G. & Reiffers, J. (2002)
Follow-up of complete cytogenetic remission in
patients with chronic myeloid leukemia after
cessation of interferon alfa. Journal of Clinical
Oncology, 20, 214220.
Mahon, F.X., Rea, D., Guilhot, J., Guilhot, F.,
Huguet, F., Nicolini, F., Legros, L., Charbonnier,
A., Guerci, A., Varet, B., Etienne, G., Reiffers, J.
& Rousselot, P. (2010) Discontinuation of
imatinib in patients with chronic myeloid
leukaemia who have maintained complete
molecular remission for at least 2 years: the pro-
spective, multicentre Stop Imatinib (STIM) trial.
Lancet Oncology, 11, 10291035.
Mahon, F.X., Nicolini, F.E., Noel, M.P., Escoffre,
M., Charbonnier, A., Rea, D., Dubruille, V.,
Varet, B.R., Legros, L., Guerci, A., Etienne, G.,
Guilhot, F., Dulucq, S., Rousselot, P. & Guilhot,
J. (2013) Preliminary report of the STIM2 study:
a multicenter stop imatinib trial for chronic
phase chronic myeloid leukemia de novo
patients on imatinib. Blood (ASH Annual Meet-
ing Abstracts), 122, 654.
Michor, F., Hughes, T.P., Iwasa, Y., Branford, S.,
Shah, N.P., Sawyers, C.L. & Nowak, M.A.
(2005) Dynamics of chronic myeloid leukaemia.
Nature, 435, 12671270.
Molldrem, J.J., Lee, P.P., Wang, C., Felio, K., Kan-
tarjian, H.M., Champlin, R.E. & Davis, M.M.
(2000) Evidence that specific T lymphocytes
may participate in the elimination of chronic
myelogenous leukemia. Nature Medicine, 6,
10181023.
Mori, S., Vagge, E., le Coutre, P., Abruzzese, E.,
Martino, B., Pungolino, E., Elena, C., Pierri, I.,
2014 John Wiley & Sons Ltd

Assouline, S., DEmilio, A., Gozzini, A., Giraldo,
P., Stagno, F., Nagler, A., Iurlo, A., Rizzo, C.,
Fossati, A., Pogliani, E.M., Kim, D.W., Pirola,
A., Piazza, R. & Gambacorti-Passerini, C. (2013)
Validation of digital-PCR analysis through pro-
grammed imatinib interruption in Q-RT-PCR
negative chronic myeloid leukemia patients.
Blood (ASH Annual Meeting Abstracts), 122,
4040.
Mughal, T.I., Yong, A., Szydlo, R.M., Dazzi, F.,
Olavarria, E., van Rhee, F., Kaeda, J., Cross,
N.C., Craddock, C., Kanfer, E., Apperley, J. &
Goldman, J.M. (2001) Molecular studies in
patients with chronic myeloid leukaemia in
remission 5 years after allogeneic stem cell
transplant define the risk of subsequent
relapse. British Journal of Haematology, 115,
569574.
Mumprecht, S., Schurch, C., Schwaller, J., Solent-
haler, M. & Ochsenbein, A.F. (2009) Pro-
grammed death 1 signaling on chronic myeloid
leukemia-specific T cells results in T-cell exhaus-
tion and disease progression. Blood, 114, 1528
1536.
Norkin, M., Uberti, J.P. & Schiffer, C.A. (2011)
Very late recurrences of leukemia: why does leu-
kemia awake after many years of dormancy?
Leukemia Research, 35, 139144.
Ohyashiki, K., Katagiri, S., Tauchi, T., Ohyashiki,
J.H., Maeda, Y., Matsumura, I. & Kyo, T.
(2012) Increased natural killer cells and
decreased CD3(+)CD8(+)CD62L(+) T cells in
CML patients who sustained complete molecular
remission after discontinuation of imatinib.
British Journal of Haematology, 157, 254256.
Preudhomme, C., Guilhot, J., Nicolini, F.E.,
Guerci-Bresler, A., Rigal-Huguet, F., Maloisel,
F., Coiteux, V., Gardembas, M., Berthou, C.,
Vekhoff, A., Rea, D., Jourdan, E., Allard, C.,
Delmer, A., Rousselot, P., Legros, L., Berger, M.,
Corm, S., Etienne, G., Roche-Lestienne, C.,
Eclache, V., Mahon, F.X. & Guilhot, F. (2010)
Imatinib plus peginterferon alfa-2a in chronic
myeloid leukemia. New England Journal of Medi-
cine, 363, 25112521.
Radich, J.P., Gehly, G., Gooley, T., Bryant, E.,
Clift, R.A., Collins, S., Edmands, S., Kirk, J.,
Lee, A., Kessler, P., Schoch, G., Buckner, C.D.,
Sullivan, K.M., Appelbaum, F.R. & Thomas,
E.D. (1995) Polymerase chain reaction detection
of the BCR-ABL fusion transcript after alloge-
neic marrow transplantation for chronic mye-
loid leukemia: results and implications in 346
patients. Blood, 85, 26322638.
Rea, D., Rousselot, P., Guilhot, F., Tulliez, M.,
Nicolini, F., Guerci-Bresler, A., Legros, L.,
Gardembas, M., Giraudier, S., Guillerm, G. &
Mahon, F. (2012) Discontinuation of second
generation (2G) tyrosine kinase inhibitors (TKI)
in chronic phase (CP)-chronic myeloid leukemia
(CML) patients with stable undetectable
2014 John Wiley & Sons Ltd
British Journal of Haematology, 2014, 166, 311
BCR-ABL transcripts. Blood (ASH Annual Meet-
ing Abstracts), 120, 916.
Rea, D., Dulphy, N., Henry, G., Guilhot, J., Guil-
hot, F., Nicolini, F.E., Legros, L., Rousselot, P.,
Mahon, F.X. & Toubert, A. (2013) Low natural
killer (NK) cell counts and functionality are
associated with molecular relapse after imatinib
discontinuation in patients (pts) with chronic
phase (CP)-chronic myeloid leukemia (CML)
with undetectable BCR-ABL transcripts for at
least 2 years: preliminary results from immuno-
stim, On Behalf Of STIM Investigators. Blood
(ASH Annual Meeting Abstracts), 122, 856.
Roeder, I., Horn, M., Glauche, I., Hochhaus, A.,
Mueller, M.C. & Loeffler, M. (2006) Dynamic
modeling of imatinib-treated chronic myeloid
leukemia: functional insights and clinical impli-
cations. Nature Medicine, 12, 11811184.
Ross, D.M., Branford, S., Seymour, J.F., Schwarer,
A.P., Arthur, C., Bartley, P.A., Slader, C., Field,
C., Dang, P., Filshie, R.J., Mills, A.K., Grigg,
A.P., Melo, J.V. & Hughes, T.P. (2010) Patients
with chronic myeloid leukemia who maintain a
complete molecular response after stopping
imatinib treatment have evidence of persistent
leukemia by DNA PCR. Leukemia, 24, 1719
1724.
Ross, D.M., Bartley, P.A., Goyne, J., Morley, A.A.,
Seymour, J.F. & Grigg, A.P. (2011) Durable
complete molecular remission of chronic mye-
loid leukemia following dasatinib cessation,
despite adverse disease features. Haematologica,
96, 17201722.
Ross, D.M., Branford, S., Seymour, J.F., Schwarer,
A.P., Arthur, C., Yeung, D.T., Dang, P., Goyne,
J.M., Slader, C., Filshie, R.J., Mills, A.K., Melo,
J.V., White, D.L., Grigg, A.P. & Hughes, T.P.
(2013) Safety and efficacy of imatinib cessation
for CML patients with stable undetectable mini-
mal residual disease: results from the TWISTER
study. Blood, 122, 515522.
Rousselot, P., Huguet, F., Rea, D., Legros, L.,
Cayuela, J.M., Maarek, O., Blanchet, O., Marit,
G., Gluckman, E., Reiffers, J., Gardembas, M. &
Mahon, F.X. (2007) Imatinib mesylate discon-
tinuation in patients with chronic myelogenous
leukemia in complete molecular remission for
more than 2 years. Blood, 109, 5860.
Rousselot, P., Charbonnier, A., Cony-Makhoul, P.,
Agape, P., Nicolini, F.E., Varet, B., Gardembas,
M., Etienne, G., Rea, D., Roy, L., Escoffre-Barbe,
M., Guerci-Bresler, A., Tulliez, M., Prost, S.,
Spentchian, M., Cayuela, J.M., Reiffers, J., Cho-
mel, J.C., Turhan, A., Guilhot, J., Guilhot, F. &
Mahon, F.X. (2014) Loss of major molecular
response as a trigger for restarting tyrosine
kinase inhibitor therapy in patients with
chronic-phase chronic myelogenous leukemia
who have stopped imatinib after durable unde-
tectable disease. Journal of Clinical Oncology, 32,
424430.
Review
Schmidt, M., Hochhaus, A., Nitsche, A., Hehl-
mann, R. & Neubauer, A. (2001) Expression of
nuclear transcription factor interferon consensus
sequence binding protein in chronic myeloid
leukemia correlates with pretreatment risk fea-
tures and cytogenetic response to interferon-
alpha. Blood, 97, 36483650.
Score, J., Calasanz, M.J., Ottman, O., Pane, F.,
Yeh, R.F., Sobrinho-Simoes, M.A., Kreil, S.,
Ward, D., Hidalgo-Curtis, C., Melo, J.V., Wie-
mels, J., Nadel, B., Cross, N.C. & Grand, F.H.
(2010) Analysis of genomic breakpoints in p190
and p210 BCR-ABL indicate distinct mecha-
nisms of formation. Leukemia, 24, 17421750.
Sokal, J.E., Cox, E.B., Baccarani, M., Tura, S., Gomez,
G.A., Robertson, J.E., Tso, C.Y., Braun, T.J., Clark-
son, B.D., Cervantes, F., Rozman, C. and The
Italian Cooperative CML Study Group. (1984)
Prognostic discrimination in good-risk chronic
granulocytic leukemia. Blood, 63, 789799.
Takahashi, N., Kyo, T., Maeda, Y., Sugihara, T.,
Usuki, K., Kawaguchi, T., Usui, N., Okamoto,
S., Ohe, Y., Ohtake, S., Kitamura, K., Yamamot-
o, M., Teshima, H., Motoji, T., Tamaki, T., Saw-
ada, K. & Ohyashiki, K. (2012) Discontinuation
of imatinib in Japanese patients with chronic
myeloid leukemia. Haematologica, 97, 903906.
Tanaka, H., Imagawa, J., Okada, M., Nakamae, H.,
Hino, M., Murai, K., Ishida, Y., Kumagai, T.,
Sato, S., Ohashi, K., Sakamaki, H., Wakita, H.,
Uoshima, N., Nakagawa, Y., Minami, Y., Oga-
sawara, M., Takeoka, T., Akasaka, H., Morita,
S., Sakamoto, J. & Kimura, S. (2013) Discontin-
uation of dasatinib in patients with CML who
have maintained complete molecular response
for at least one year: results from a prospective
discontinuation (DADI) trial. Blood (ASH
Annual Meeting Abstracts), 122, 3998.
Verma, D., Kantarjian, H., Jain, N. & Cortes, J.
(2008) Sustained complete molecular response
after imatinib discontinuation in a patient with
chronic myeloid leukemia not previously
exposed to interferon alpha. Leukaemia & Lym-
phoma, 49, 13991402.
Warsch, W., Walz, C. & Sexl, V. (2013) JAK of all
trades: JAK2-STAT5 as novel therapeutic targets
in BCR-ABL1+ chronic myeloid leukemia.
Blood, 122, 21672175.
Yhim, H.Y., Lee, N.R., Song, E.K., Yim, C.Y., Jeon,
S.Y., Shin, S., Kim, J.A., Kim, H.S., Cho, E.H. &
Kwak, J.Y. (2012) Imatinib mesylate discontinu-
ation in patients with chronic myeloid leukemia
who have received front-line imatinib mesylate
therapy and achieved complete molecular
response. Leukemia Research, 36, 689693.
Zhang, B., Irvine, D., Ho, Y.W., Buonamici, S., Man-
ley, P., Holyoake, T.L., Copland, M. & Bhatia, R.
(2010) Inhibition of chronic myeloid leukemia
stem cells by the combination of the hedgehog
pathway inhibitor LDE225 with nilotinib. Blood
(ASH Annual Meeting Abstracts), 116, 514.
11