Management of polycythaemia vera: a critical review of current data Mary F. McMullin, 1 Bridget S. Wilkins 2 and Claire N. Harrison 2 1 Centre for Cancer Research and Cell Biology, Queen’s University, Belfast, and 2 Departments of Cellular Pathology and Haematology, Guy’s and St Thomas’ NHS Foundation Trust, London, UK Summary Polycythaemia vera (PV) is a chronic blood cancer; its clini- cal features are dominated by myeloproliferation (erythrocy- tosis, often leucocytosis and/or thrombocytosis) and a tendency for thrombosis and transformation to myelofibrosis or acute myeloid leukaemia. In the past 10 years the patho- physiology of this condition has been defined as JAK/STAT pathway activation, almost always due to mutations in JAK2 exons 12 or 14 (JAK2 V617F). In the same time period our understanding of the optimal management of PV has expanded, most recently culminating in the approval of JAK inhibitors for the treatment of PV patients who are resistant or intolerant to therapy with hydroxycarbamide. It has also been demonstrated that life expectancy for many patients with PV is not normal, nor is their quality of life. We criti- cally explore these findings and discuss their impact. In addi- tion, we highlight persisting gaps in our current management strategy; for example, what is the optimal first line cytore- ductive therapy and, indeed, which patients need cytoreduc- tive drugs. Keywords: polycythaemia vera, JAK2, myeloproliferative. Louis Henri Vaquez (1860–1936), a French physician, was the first to describe polycythaemia vera (PV) in 1892. Vaquez also speculated that the increased red blood cell count was due to haemopoietic hyperactivity (Vaquez, 1892). However the condition is still sometimes referred to as Osler-Vaquez disease, recognizing that William Osler (1849–1919) also described an early case. However, it was not until a seminal paper in 1951, in which Dameshek identified the features of what he called ‘myeloproliferative disorders’ (MPD), that this term was used for what he described as PV, myelofibrosis and chronic myeloid leukaemia (Dameshek, 1951). He also speculated that these disorders were ‘all somewhat variable manifestations of proliferative activity of the bone marrow cells perhaps due to an undiscovered stimulus’ (Dameshek, 1951). From that time onwards, the clinical and diagnostic characteristics of PV were defined. Patients were identified as having an increased red cell mass (RCM) and combinations of splenomegaly, genetic markers of clonal haemopoiesis, thrombocytosis, neutrophil leucocytosis, endogenous ery- throid colony growth and reduced serum erythropoietin (Pearson & Messinezy, 1996). In 2005, a number of groups identified a single mutation, JAK2 V617F, in the majority of patients with PV (Baxter et al, 2005; James et al, 2005; Kralovics et al, 2005; Levine et al, 2005). Soon after this, mutations were identified in exon 12 of JAK2 in many of the minority of PV patients who did not have a JAK2 V617F mutation (Scott et al, 2007). These critical discoveries meant not only that it was possible to demonstrate a clonal marker in the vast majority of patients with PV but also that our diagnostic criteria for PV could be more precisely defined. It also became apparent that the JAK2 V617F mutation was also seen in the other myeloproliferative neoplasms (MPN), i.e., in primary myelofibrosis and essential thrombocythaemia (ET) as well as occasionally in a few patients with other chronic myeloid neoplasms. This has provoked research to define how one mutation can lead to several different disease phenotypes and also a discussion as to whether the MPN were not separate entities but rather formed a biological con- tinuum (Campbell et al, 2005). Lastly, discovery of the JAK2 V617F mutation also led to a more intense focus on MPN in general, resulting in their redesignation as true neoplasms and the very recent development of new therapeutic options. In addition to advances driven by molecular discoveries, data from clinical trials has been invaluable for informing our management approach for PV. In the past, such studies have involved small numbers of patients and most have been retrospective. This has left areas of uncertainty and inconsis- tency in clinical guidance. For example, Pearson and Wether- ley-Mein (1978) demonstrated a correlation between haematocrit (Hct) and cerebral blood flow as well as throm- botic events, leading to the recommended target Hct of 0 45 for treatment. This has been re-evaluated only recently in a randomized trial, confirming both the historical data and the validity of the target Hct (Marchioli et al, 2013). Similarly, Correspondence: Professor Claire N. Harrison, Department of Haematology, Guy’s and St Thomas’ NHS Foundation Trust, St Thomas’ Hospita, Lambeth Palace Road, London SE1 7EH, UK. E-mail: [email protected] review ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349 First published online 22 October 2015 doi: 10.1111/bjh.13812
Management of polycythaemia vera: a critical review of current data
Apr 03, 2023
Polycythaemia vera (PV) is a chronic blood cancer; its clinical features are dominated by myeloproliferation (erythrocytosis, often leucocytosis and/or thrombocytosis) and a tendency for thrombosis and transformation to myelofibrosis or acute myeloid leukaemia. In the past 10 years the pathophysiology of this condition has been defined as JAK/STAT pathway activation, almost always due to mutations in JAK2 exons 12 or 14 (JAK2 V617F). In the same time period our understanding of the optimal management of PV has expanded, most recently culminating in the approval of JAK inhibitors for the treatment of PV patients who are resistant or intolerant to therapy with hydroxycarbamide.
Welcome message from author
In 2008, the World Health Organization (WHO) revised the
diagnostic criteria for PV, incorporating JAK2 V617F and
JAK2 exon 12 data (Vardiman et al, 2009). Major criteria
became haemoglobin (Hb) >185 g/l in males or >165 g/l in
females, or evidence of an increased RCM, and a V617F or
exon 12 JAK2 mutation. Minor criteria were defined as a
bone marrow trephine biopsy specimen showing hypercellularity with panmyelosis, subnormal erythropoietin levels and
endogenous erythroid colony growth. To make a diagnosis of
PV, fulfilment of both major and one minor criteria or the
first major and two minor criteria where required
Transcript
Management of polycythaemia vera: a critical review of current dataManagement of polycythaemia vera: a critical review of current data
Mary F. McMullin,1 Bridget S. Wilkins2 and Claire N. Harrison2
1Centre for Cancer Research and Cell Biology, Queen’s University, Belfast, and 2Departments of Cellular Pathology and Haematology,
Guy’s and St Thomas’ NHS Foundation Trust, London, UK
Summary
Polycythaemia vera (PV) is a chronic blood cancer; its clini-
cal features are dominated by myeloproliferation (erythrocy-
tosis, often leucocytosis and/or thrombocytosis) and a
tendency for thrombosis and transformation to myelofibrosis
or acute myeloid leukaemia. In the past 10 years the patho-
physiology of this condition has been defined as JAK/STAT
pathway activation, almost always due to mutations in JAK2
exons 12 or 14 (JAK2 V617F). In the same time period our
understanding of the optimal management of PV has
expanded, most recently culminating in the approval of JAK
inhibitors for the treatment of PV patients who are resistant
or intolerant to therapy with hydroxycarbamide. It has also
been demonstrated that life expectancy for many patients
with PV is not normal, nor is their quality of life. We criti-
cally explore these findings and discuss their impact. In addi-
tion, we highlight persisting gaps in our current management
strategy; for example, what is the optimal first line cytore-
ductive therapy and, indeed, which patients need cytoreduc-
tive drugs.
Louis Henri Vaquez (1860–1936), a French physician, was
the first to describe polycythaemia vera (PV) in 1892. Vaquez
also speculated that the increased red blood cell count was
due to haemopoietic hyperactivity (Vaquez, 1892). However
the condition is still sometimes referred to as Osler-Vaquez
disease, recognizing that William Osler (1849–1919) also
described an early case. However, it was not until a seminal
paper in 1951, in which Dameshek identified the features of
what he called ‘myeloproliferative disorders’ (MPD), that this
term was used for what he described as PV, myelofibrosis
and chronic myeloid leukaemia (Dameshek, 1951). He also
speculated that these disorders were ‘all somewhat variable
manifestations of proliferative activity of the bone marrow
cells perhaps due to an undiscovered stimulus’ (Dameshek,
1951). From that time onwards, the clinical and diagnostic
characteristics of PV were defined. Patients were identified as
having an increased red cell mass (RCM) and combinations
of splenomegaly, genetic markers of clonal haemopoiesis,
thrombocytosis, neutrophil leucocytosis, endogenous ery-
throid colony growth and reduced serum erythropoietin
(Pearson & Messinezy, 1996). In 2005, a number of groups
identified a single mutation, JAK2 V617F, in the majority of
patients with PV (Baxter et al, 2005; James et al, 2005;
Kralovics et al, 2005; Levine et al, 2005). Soon after this,
mutations were identified in exon 12 of JAK2 in many of the
minority of PV patients who did not have a JAK2 V617F
mutation (Scott et al, 2007). These critical discoveries meant
not only that it was possible to demonstrate a clonal marker
in the vast majority of patients with PV but also that our
diagnostic criteria for PV could be more precisely defined. It
also became apparent that the JAK2 V617F mutation was
also seen in the other myeloproliferative neoplasms (MPN),
i.e., in primary myelofibrosis and essential thrombocythaemia
(ET) as well as occasionally in a few patients with other
chronic myeloid neoplasms. This has provoked research to
define how one mutation can lead to several different disease
phenotypes and also a discussion as to whether the MPN
were not separate entities but rather formed a biological con-
tinuum (Campbell et al, 2005). Lastly, discovery of the JAK2
V617F mutation also led to a more intense focus on MPN in
general, resulting in their redesignation as true neoplasms
and the very recent development of new therapeutic options.
In addition to advances driven by molecular discoveries,
data from clinical trials has been invaluable for informing
our management approach for PV. In the past, such studies
have involved small numbers of patients and most have been
retrospective. This has left areas of uncertainty and inconsis-
tency in clinical guidance. For example, Pearson and Wether-
ley-Mein (1978) demonstrated a correlation between
haematocrit (Hct) and cerebral blood flow as well as throm-
botic events, leading to the recommended target Hct of 045 for treatment. This has been re-evaluated only recently in a
randomized trial, confirming both the historical data and the
validity of the target Hct (Marchioli et al, 2013). Similarly,
Correspondence: Professor Claire N. Harrison, Department of
Haematology, Guy’s and St Thomas’ NHS Foundation Trust, St
Thomas’ Hospita, Lambeth Palace Road, London SE1 7EH, UK.
E-mail: [email protected]
review
ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
First published online 22 October 2015 doi: 10.1111/bjh.13812
controlled trial data supporting the benefit of low-dose
aspirin in patients with PV only emerged, just prior to the
description of JAK2 V617F, from the European Collaboration
on Low-Dose Aspirin in Polycythemia Vera (ECLAP) study
(Landolfi et al, 2004). In these examples, recent data clarified
and validated standard practice; however, points of contro-
versy remain and new controversies have developed.
In this article, we review the current management of PV
and provide an algorithm for management (Fig 1) which is
very similar to that developed through the British Committee
for Standards in Haematology (BCSH) process (McMullin
et al, 2005). Furthermore, we critically evaluate new data and
discuss points of controversy; for example, standard diagnos-
tic criteria including recent proposals to revise these, and the
utility of current criteria to describe transformation from PV
to post-PV myelofibrosis (PPV-MF). Problems for patient
management reside mainly in risk stratification to decide
who needs cytoreductive therapy, which cytoreductive ther-
apy to use, what the treatment target is, the measurement
and documentation of symptoms, and their management
(especially pruritus). Finally, we discuss how to assess and
make best clinical use of data suggesting benefit for novel
therapies.
diagnostic criteria for PV, incorporating JAK2 V617F and
JAK2 exon 12 data (Vardiman et al, 2009). Major criteria
became haemoglobin (Hb) >185 g/l in males or >165 g/l in
females, or evidence of an increased RCM, and a V617F or
exon 12 JAK2 mutation. Minor criteria were defined as a
bone marrow trephine biopsy specimen showing hypercellu-
larity with panmyelosis, subnormal erythropoietin levels and
endogenous erythroid colony growth. To make a diagnosis of
PV, fulfilment of both major and one minor criteria or the
first major and two minor criteria where required (Table I).
These criteria simplified what had been a highly complex
diagnostic pathway, but emphasized the importance of a
bone marrow examination. Characteristic features of the
bone marrow in PV are illustrated in Fig 2A, B. In contrast,
the BCSH guideline group (also using JAK2 mutation data)
concluded that a diagnosis of PV could be made if the Hct
was >052 in males or >048 in females or, alternatively, a
RCM above 125% of predicted, plus JAK2 mutation was pre-
sent (Table I): the evidence for these thresholds of Hct and
RCM is reviewed in the BCSH guideline manuscript
Fig 1. Algorithm for proposed management of
polycythaemia vera (each facet requires evalua-
tion at patient visits). Hct, haematocrit; IFN,
interferon; HC, hydroxycarbamide; PUVA,
psoralen (P) and ultraviolet A (UVA) therapy;
MPN, myeloproliferative neoplasm. *Suggest use of a validated tool such as MPN symptom
assessment form. **Unless contraindicated, caution if platelet count elevated. ***A lower
Hct target is sometimes required for symp-
tomatic patients.
Review
338 ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
(McMullin et al, 2007). These criteria use Hct rather than
Hb as a surrogate marker of increased RCM and do not
require a bone marrow examination in those for whom a
clonal marker is shown. They also continue to provide a
means for diagnosing PV, using a robust combination of dif-
ferent factors, for the ~1% of PV patients lacking a detectable
mutation in exon 12 or 14 of JAK2. Operationally, the BCSH
criteria for PV have been shown recently to perform better
than those from the WHO (Alvarez-Larran et al, 2014a,
2015; Barbui et al, 2014a) with better rates of both sensitivity
and specificity.
As discussed above, the diagnosis of PV requires an increased
RCM or the use of a surrogate marker, such as Hb or Hct.
However, it has been shown that many patients with
increased RCM do not have an increased Hb (Johansson
et al, 2005). A separate study investigated a cohort of 397
patients with a JAK2 mutation and a bone marrow that was
said to show PV morphology (Barbui et al, 2014a). Of this
group, 35% did not have Hb sufficiently elevated to fulfil the
WHO diagnostic criteria of PV. Using the BCSH Hct criteria,
this non-diagnostic cohort fell to 15%; suggesting that Hct is
a better indicator of raised RCM than Hb. However, this still
left a considerable population who did not achieve an appro-
priate diagnosis of PV. Such patients have been described as
having ‘masked PV’. Further analysis of the masked PV
group showed that they had an inferior overall survival com-
pared to the overt PV group (Barbui et al, 2014a,b)
although, in contrast, Spanish researchers found that out-
comes were indistinguishable (Alvarez-Larran et al, 2015).
The reasons for the suggested inferior survival are unclear
but a possible interpretation is that patients were not being
treated to lower their Hct sufficiently. It is unclear whether
patients in the ‘masked PV’ category had iron deficiency and
whether attention to this would have hastened or eased their
diagnosis.
nosis of PV supported by bone marrow histology and the
ability to discriminate the bone marrow appearances of PV
from those of other MPN, in particular ET, an illustrative
case is shown in Fig 3. In these studies, all bone marrow his-
tology was reviewed by a single expert pathologist who was
blinded to outcome but not to the Hb (Barbui et al, 2014a,
b). In a Danish study (Madelung et al, 2013), seven experi-
enced haematopathologists reviewed a large number of bone
marrow trephine specimens from MPN patients and controls,
blinded to all clinical data except age. Histological consensus
with the clinical diagnosis in this study was 71%, demon-
strating that it can be difficult to make a definitive diagnosis
on the basis of histology alone, even for experienced individ-
uals in the wider haematology and haematopathology com-
munity (Madelung et al, 2013), reflecting real-life challenges
in clinical practice.
Using data from the PV and masked PV studies, an analy-
sis of a cohort of ET patients suggested that Hb of 165 g/l in
males and 160 g/l in females, or a Hct of 049 in males and
048 in females provide the best discrimination between PV
and ET (Barbui et al, 2014c). It is important to note that
these values are close to, or within, the normal range, and
are being used here to discriminate PV from ET rather than
PV from normal individuals.
In response to the issues of a difficulty in defining a cut-
off Hb or Hct that is more sensitive and specific for a diag-
nosis of PV in patients with suspected PV and in patients
with raised platelet counts who might have PV rather than
ET, revised WHO criteria are now being proposed. These
Table I. Diagnostic criteria.
(Barbui et al, 2014c)
or other evidence of increased red cell
volume
OR
predicted)
Hct >049 (male), >048 (female)
JAK2 V617F or other functionally similar
mutation such as JAK2 exon 12 mutation
Mutation in JAK2 Bone marrow: trilineage myeloproliferation
with pleomorphic megakaroyocytes
panmyelosis
criterion or first major criterion plus two
minor criteria
Both criteria present All three major or first two major and one
minor
WHO, World Health Organization; BCSH, British Committee for Standards in Haematology; EPO, erythropoietin.
Review
ª 2015 John Wiley & Sons Ltd 339 British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
icense
represent three new major criteria; (i) Hb >165 g/l in males
or >160 g/l in females (or Hct >049 in males or >048 in
females), (ii) trilineage myeloproliferation with pleomorphic
megakaryocytes and (iii) presence of a JAK2 mutation. A
suboptimal erythropoetin level would constitute a new minor
criterion. All three major criteria, or the first two major cri-
teria plus the minor criterion (Table I) are required for diag-
nosis. Application of these criteria will create the possibility
of a PV diagnosis for individuals with Hb well within the
normal range (Barbui et al, 2014c). They will also mandate
bone marrow examination for diagnosis in all patients,
regardless of whether or not clonal disease is shown to be
present. Our opinion is that these criteria are problematic
and that a more practical approach to this dilemma might
be to emphasize the need to exclude PV carefully in patients
for whom a diagnosis of ET is being considered. The revised,
more stringent Hb/Hct thresholds may be useful to trigger a
more critical consideration of possible PV in this specific sit-
uation. Residual challenges would be the need for a high
standard of diagnostic acumen to discriminate these condi-
tions through their histological characteristics, and/or the
availability of RCM measurement to document an absolute
erythocytosis.
Transformation to PPV-MF occurs in approximately 25% of
patients with PV over time and, given the reduced life expec-
tancy for patients in whom this occurs, consideration of
curative treatment with allogeneic stem cell transplantation
and the utility of therapies such a JAK inhibitors (Cervantes,
2014) make this diagnosis important to establish appropri-
ately. Current criteria for PPV-MF were developed in by Bar-
osi et al (2008a,b), as shown in Table II. In some
circumstances, application of these criteria can be challenging
because key features may develop slowly. In particular, the
requirement for a relatively high degree of reticulin fibrosis
(at least grade 2 on a 3-point scale, such as that of the
WHO, and grade 3 on a 4-point scale, such as Bauermeister)
can exclude some patients who might benefit in some cir-
cumstances from specific interventions in a variety of clinical
trials (Gowin et al, 2015). Figure 4A, B illustrates trephine
biopsy features in PPV-MF.
Aims of treatment in PV
The natural history of PV is long, with survival usually less
than, but comparable to, the general population (Hultcrantz
et al, 2012; Tefferi et al, 2013). However, in patients with
PV, events may occur that cause morbidity and limit life
expectancy. This is well demonstrated in a historic retrospec-
tive study in which survival was investigated in a group of
(A)
(B)
and pleomorphic, clustered megakayrocytes. Haematoxylin and Eosin
(H&E) stain; original magnification 940. (B) Borderline increase in
reticulin fibres (World Health Organization fibrosis score 0–1) accompanying panmyelosis. Gomori silver stain for reticulin; original
magnification 940.
Fig 3. ‘Masked PV’. A 56-year-old female with a 4-year history of
thrombocytosis. Previous bone marrow trephine samples reported as
‘essential thrombocythaemia-like’ at initial presentation, with lower
cellularity but noted pleomorphism among megakaryocytes. This
sample is hypercellular due to panmyelosis, with marked pleomor-
phism and clustering of prominent megakaryocytes. H&E stain; orig-
inal magnification 920.
Review
340 ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
survived at 18 months from first presentation and the main
cause of death was thrombosis (Chievitz & Thiede, 1962).
However, in those who survive longer, transformation to
myelofibrosis and acute leukaemia are life-limiting events
(Kiladjian et al, 2011).
In treating PV, our aim is to reduce complications and
thus improve survival. Thromboembolic events are the major
cause of morbidity and mortality; consequently, their preven-
tion is the main aim of any treatment strategy. A difficulty
in assessing the ability of novel therapies to address this is
the lack of an appropriate surrogate endpoint for the mea-
surement of prevention of thromboembolic complications
and the low prevalence of these complications in patients
treated with current therapies. However we would wish to
emphasize that rates of thromboembolic events remain
higher than the normal population and are an appropriate
target. Prevention of progression to myelofibrosis or acute
leukaemia, and avoiding administration of any treatment that
would increase disease-related tendency to progression would
also be desirable. Fulfilment of these aims is presently an
aspiration, as it would require treatments with demonstrated
ability to prevent progression and no such therapies have as
yet been discovered. Nonetheless, it remains a valuable aspi-
ration when evaluating novel therapies, as these targets are
valid and worthy of perusal.
While the main aim of therapy in PV is directed at the
prevention of thromboembolic events, in the past decade
there has been an increasing appreciation of the additional
individual and societal burdens associated with impaired
quality of life in PV patients, manifest not only by sequelae
of thrombotic events but also by disease-related symptoms.
Large studies have demonstrated a high prevalence of symp-
toms and impaired quality of life in specific patient cohorts.
These studies have also established and validated tools for
the quantitation of such effects (Geyer & Mesa, 2014; Scher-
ber et al, 2014). To what extent these symptoms will respond
to standard therapies and differ from symptoms reported by
case-matched controls remains to be assessed formally and is
important for the field. Nonetheless, it is clear that some
patients have a high burden of symptoms (often pruritus)
that are clearly related to disease and not ameliorated by
standard treatments. These issues have been brought more
into focus by the development of JAK inhibitors, one benefit
of which appears to be to successful control of disease-related
symptoms. This is discussed in more detail below but the rel-
evance of symptom control as an aim of therapy merits its
inclusion as a separate item for consideration here.
Risk groups: what is high-risk PV?
Treatment aimed at the prevention of thromboembolic
events can involve measures such as venesection to a target
Hct of 045 (rarely, lower targets are used in symptomatic
Table II. Criteria for the diagnosis of post-polycythaemia vera
myelofibrosis (PPV-MF) (Barosi et al, 2008a,b).
Major criteria – both required 1 Documentation of a previous diagnosis of polycythaemia vera
as defined by the World Health Organization criteria
2 Bone marrow fibrosis grade 2–3 (on 0–3 scale) or grade 3–4 (on 0–4 scale)
Minor criteria – at least two required 1 Anaemia or sustained loss of requirement of either phlebotomy
(in the absence of cytoreductive therapy) or cytoreductive
treatment for erythrocytosis
3 Increasing splenomegaly, defined as either an increase in
palpable splenomegaly of ≥5 cm (distance of the tip of the
spleen from the left costal margin) or the appearance of a
newly palpable splenomegaly
(>375°C)
A patient with longstanding polycythaemia vera, now thrombocy-
topenic, unresponsive to hydroxycarbamide and anagrelide. Biopsy is
hypercellular with obvious collagen fibrosis. At this stage the histo-
logical picture is indistinguishable from primary myelofibrosis. H&E
stain; original magnification 920. (B) PPV-MF. Gomori reticulin
stain to accompany H&E shown in panel A; original magnification
920.
Review
ª 2015 John Wiley & Sons Ltd 341 British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions…
Mary F. McMullin,1 Bridget S. Wilkins2 and Claire N. Harrison2
1Centre for Cancer Research and Cell Biology, Queen’s University, Belfast, and 2Departments of Cellular Pathology and Haematology,
Guy’s and St Thomas’ NHS Foundation Trust, London, UK
Summary
Polycythaemia vera (PV) is a chronic blood cancer; its clini-
cal features are dominated by myeloproliferation (erythrocy-
tosis, often leucocytosis and/or thrombocytosis) and a
tendency for thrombosis and transformation to myelofibrosis
or acute myeloid leukaemia. In the past 10 years the patho-
physiology of this condition has been defined as JAK/STAT
pathway activation, almost always due to mutations in JAK2
exons 12 or 14 (JAK2 V617F). In the same time period our
understanding of the optimal management of PV has
expanded, most recently culminating in the approval of JAK
inhibitors for the treatment of PV patients who are resistant
or intolerant to therapy with hydroxycarbamide. It has also
been demonstrated that life expectancy for many patients
with PV is not normal, nor is their quality of life. We criti-
cally explore these findings and discuss their impact. In addi-
tion, we highlight persisting gaps in our current management
strategy; for example, what is the optimal first line cytore-
ductive therapy and, indeed, which patients need cytoreduc-
tive drugs.
Louis Henri Vaquez (1860–1936), a French physician, was
the first to describe polycythaemia vera (PV) in 1892. Vaquez
also speculated that the increased red blood cell count was
due to haemopoietic hyperactivity (Vaquez, 1892). However
the condition is still sometimes referred to as Osler-Vaquez
disease, recognizing that William Osler (1849–1919) also
described an early case. However, it was not until a seminal
paper in 1951, in which Dameshek identified the features of
what he called ‘myeloproliferative disorders’ (MPD), that this
term was used for what he described as PV, myelofibrosis
and chronic myeloid leukaemia (Dameshek, 1951). He also
speculated that these disorders were ‘all somewhat variable
manifestations of proliferative activity of the bone marrow
cells perhaps due to an undiscovered stimulus’ (Dameshek,
1951). From that time onwards, the clinical and diagnostic
characteristics of PV were defined. Patients were identified as
having an increased red cell mass (RCM) and combinations
of splenomegaly, genetic markers of clonal haemopoiesis,
thrombocytosis, neutrophil leucocytosis, endogenous ery-
throid colony growth and reduced serum erythropoietin
(Pearson & Messinezy, 1996). In 2005, a number of groups
identified a single mutation, JAK2 V617F, in the majority of
patients with PV (Baxter et al, 2005; James et al, 2005;
Kralovics et al, 2005; Levine et al, 2005). Soon after this,
mutations were identified in exon 12 of JAK2 in many of the
minority of PV patients who did not have a JAK2 V617F
mutation (Scott et al, 2007). These critical discoveries meant
not only that it was possible to demonstrate a clonal marker
in the vast majority of patients with PV but also that our
diagnostic criteria for PV could be more precisely defined. It
also became apparent that the JAK2 V617F mutation was
also seen in the other myeloproliferative neoplasms (MPN),
i.e., in primary myelofibrosis and essential thrombocythaemia
(ET) as well as occasionally in a few patients with other
chronic myeloid neoplasms. This has provoked research to
define how one mutation can lead to several different disease
phenotypes and also a discussion as to whether the MPN
were not separate entities but rather formed a biological con-
tinuum (Campbell et al, 2005). Lastly, discovery of the JAK2
V617F mutation also led to a more intense focus on MPN in
general, resulting in their redesignation as true neoplasms
and the very recent development of new therapeutic options.
In addition to advances driven by molecular discoveries,
data from clinical trials has been invaluable for informing
our management approach for PV. In the past, such studies
have involved small numbers of patients and most have been
retrospective. This has left areas of uncertainty and inconsis-
tency in clinical guidance. For example, Pearson and Wether-
ley-Mein (1978) demonstrated a correlation between
haematocrit (Hct) and cerebral blood flow as well as throm-
botic events, leading to the recommended target Hct of 045 for treatment. This has been re-evaluated only recently in a
randomized trial, confirming both the historical data and the
validity of the target Hct (Marchioli et al, 2013). Similarly,
Correspondence: Professor Claire N. Harrison, Department of
Haematology, Guy’s and St Thomas’ NHS Foundation Trust, St
Thomas’ Hospita, Lambeth Palace Road, London SE1 7EH, UK.
E-mail: [email protected]
review
ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
First published online 22 October 2015 doi: 10.1111/bjh.13812
controlled trial data supporting the benefit of low-dose
aspirin in patients with PV only emerged, just prior to the
description of JAK2 V617F, from the European Collaboration
on Low-Dose Aspirin in Polycythemia Vera (ECLAP) study
(Landolfi et al, 2004). In these examples, recent data clarified
and validated standard practice; however, points of contro-
versy remain and new controversies have developed.
In this article, we review the current management of PV
and provide an algorithm for management (Fig 1) which is
very similar to that developed through the British Committee
for Standards in Haematology (BCSH) process (McMullin
et al, 2005). Furthermore, we critically evaluate new data and
discuss points of controversy; for example, standard diagnos-
tic criteria including recent proposals to revise these, and the
utility of current criteria to describe transformation from PV
to post-PV myelofibrosis (PPV-MF). Problems for patient
management reside mainly in risk stratification to decide
who needs cytoreductive therapy, which cytoreductive ther-
apy to use, what the treatment target is, the measurement
and documentation of symptoms, and their management
(especially pruritus). Finally, we discuss how to assess and
make best clinical use of data suggesting benefit for novel
therapies.
diagnostic criteria for PV, incorporating JAK2 V617F and
JAK2 exon 12 data (Vardiman et al, 2009). Major criteria
became haemoglobin (Hb) >185 g/l in males or >165 g/l in
females, or evidence of an increased RCM, and a V617F or
exon 12 JAK2 mutation. Minor criteria were defined as a
bone marrow trephine biopsy specimen showing hypercellu-
larity with panmyelosis, subnormal erythropoietin levels and
endogenous erythroid colony growth. To make a diagnosis of
PV, fulfilment of both major and one minor criteria or the
first major and two minor criteria where required (Table I).
These criteria simplified what had been a highly complex
diagnostic pathway, but emphasized the importance of a
bone marrow examination. Characteristic features of the
bone marrow in PV are illustrated in Fig 2A, B. In contrast,
the BCSH guideline group (also using JAK2 mutation data)
concluded that a diagnosis of PV could be made if the Hct
was >052 in males or >048 in females or, alternatively, a
RCM above 125% of predicted, plus JAK2 mutation was pre-
sent (Table I): the evidence for these thresholds of Hct and
RCM is reviewed in the BCSH guideline manuscript
Fig 1. Algorithm for proposed management of
polycythaemia vera (each facet requires evalua-
tion at patient visits). Hct, haematocrit; IFN,
interferon; HC, hydroxycarbamide; PUVA,
psoralen (P) and ultraviolet A (UVA) therapy;
MPN, myeloproliferative neoplasm. *Suggest use of a validated tool such as MPN symptom
assessment form. **Unless contraindicated, caution if platelet count elevated. ***A lower
Hct target is sometimes required for symp-
tomatic patients.
Review
338 ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
(McMullin et al, 2007). These criteria use Hct rather than
Hb as a surrogate marker of increased RCM and do not
require a bone marrow examination in those for whom a
clonal marker is shown. They also continue to provide a
means for diagnosing PV, using a robust combination of dif-
ferent factors, for the ~1% of PV patients lacking a detectable
mutation in exon 12 or 14 of JAK2. Operationally, the BCSH
criteria for PV have been shown recently to perform better
than those from the WHO (Alvarez-Larran et al, 2014a,
2015; Barbui et al, 2014a) with better rates of both sensitivity
and specificity.
As discussed above, the diagnosis of PV requires an increased
RCM or the use of a surrogate marker, such as Hb or Hct.
However, it has been shown that many patients with
increased RCM do not have an increased Hb (Johansson
et al, 2005). A separate study investigated a cohort of 397
patients with a JAK2 mutation and a bone marrow that was
said to show PV morphology (Barbui et al, 2014a). Of this
group, 35% did not have Hb sufficiently elevated to fulfil the
WHO diagnostic criteria of PV. Using the BCSH Hct criteria,
this non-diagnostic cohort fell to 15%; suggesting that Hct is
a better indicator of raised RCM than Hb. However, this still
left a considerable population who did not achieve an appro-
priate diagnosis of PV. Such patients have been described as
having ‘masked PV’. Further analysis of the masked PV
group showed that they had an inferior overall survival com-
pared to the overt PV group (Barbui et al, 2014a,b)
although, in contrast, Spanish researchers found that out-
comes were indistinguishable (Alvarez-Larran et al, 2015).
The reasons for the suggested inferior survival are unclear
but a possible interpretation is that patients were not being
treated to lower their Hct sufficiently. It is unclear whether
patients in the ‘masked PV’ category had iron deficiency and
whether attention to this would have hastened or eased their
diagnosis.
nosis of PV supported by bone marrow histology and the
ability to discriminate the bone marrow appearances of PV
from those of other MPN, in particular ET, an illustrative
case is shown in Fig 3. In these studies, all bone marrow his-
tology was reviewed by a single expert pathologist who was
blinded to outcome but not to the Hb (Barbui et al, 2014a,
b). In a Danish study (Madelung et al, 2013), seven experi-
enced haematopathologists reviewed a large number of bone
marrow trephine specimens from MPN patients and controls,
blinded to all clinical data except age. Histological consensus
with the clinical diagnosis in this study was 71%, demon-
strating that it can be difficult to make a definitive diagnosis
on the basis of histology alone, even for experienced individ-
uals in the wider haematology and haematopathology com-
munity (Madelung et al, 2013), reflecting real-life challenges
in clinical practice.
Using data from the PV and masked PV studies, an analy-
sis of a cohort of ET patients suggested that Hb of 165 g/l in
males and 160 g/l in females, or a Hct of 049 in males and
048 in females provide the best discrimination between PV
and ET (Barbui et al, 2014c). It is important to note that
these values are close to, or within, the normal range, and
are being used here to discriminate PV from ET rather than
PV from normal individuals.
In response to the issues of a difficulty in defining a cut-
off Hb or Hct that is more sensitive and specific for a diag-
nosis of PV in patients with suspected PV and in patients
with raised platelet counts who might have PV rather than
ET, revised WHO criteria are now being proposed. These
Table I. Diagnostic criteria.
(Barbui et al, 2014c)
or other evidence of increased red cell
volume
OR
predicted)
Hct >049 (male), >048 (female)
JAK2 V617F or other functionally similar
mutation such as JAK2 exon 12 mutation
Mutation in JAK2 Bone marrow: trilineage myeloproliferation
with pleomorphic megakaroyocytes
panmyelosis
criterion or first major criterion plus two
minor criteria
Both criteria present All three major or first two major and one
minor
WHO, World Health Organization; BCSH, British Committee for Standards in Haematology; EPO, erythropoietin.
Review
ª 2015 John Wiley & Sons Ltd 339 British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
icense
represent three new major criteria; (i) Hb >165 g/l in males
or >160 g/l in females (or Hct >049 in males or >048 in
females), (ii) trilineage myeloproliferation with pleomorphic
megakaryocytes and (iii) presence of a JAK2 mutation. A
suboptimal erythropoetin level would constitute a new minor
criterion. All three major criteria, or the first two major cri-
teria plus the minor criterion (Table I) are required for diag-
nosis. Application of these criteria will create the possibility
of a PV diagnosis for individuals with Hb well within the
normal range (Barbui et al, 2014c). They will also mandate
bone marrow examination for diagnosis in all patients,
regardless of whether or not clonal disease is shown to be
present. Our opinion is that these criteria are problematic
and that a more practical approach to this dilemma might
be to emphasize the need to exclude PV carefully in patients
for whom a diagnosis of ET is being considered. The revised,
more stringent Hb/Hct thresholds may be useful to trigger a
more critical consideration of possible PV in this specific sit-
uation. Residual challenges would be the need for a high
standard of diagnostic acumen to discriminate these condi-
tions through their histological characteristics, and/or the
availability of RCM measurement to document an absolute
erythocytosis.
Transformation to PPV-MF occurs in approximately 25% of
patients with PV over time and, given the reduced life expec-
tancy for patients in whom this occurs, consideration of
curative treatment with allogeneic stem cell transplantation
and the utility of therapies such a JAK inhibitors (Cervantes,
2014) make this diagnosis important to establish appropri-
ately. Current criteria for PPV-MF were developed in by Bar-
osi et al (2008a,b), as shown in Table II. In some
circumstances, application of these criteria can be challenging
because key features may develop slowly. In particular, the
requirement for a relatively high degree of reticulin fibrosis
(at least grade 2 on a 3-point scale, such as that of the
WHO, and grade 3 on a 4-point scale, such as Bauermeister)
can exclude some patients who might benefit in some cir-
cumstances from specific interventions in a variety of clinical
trials (Gowin et al, 2015). Figure 4A, B illustrates trephine
biopsy features in PPV-MF.
Aims of treatment in PV
The natural history of PV is long, with survival usually less
than, but comparable to, the general population (Hultcrantz
et al, 2012; Tefferi et al, 2013). However, in patients with
PV, events may occur that cause morbidity and limit life
expectancy. This is well demonstrated in a historic retrospec-
tive study in which survival was investigated in a group of
(A)
(B)
and pleomorphic, clustered megakayrocytes. Haematoxylin and Eosin
(H&E) stain; original magnification 940. (B) Borderline increase in
reticulin fibres (World Health Organization fibrosis score 0–1) accompanying panmyelosis. Gomori silver stain for reticulin; original
magnification 940.
Fig 3. ‘Masked PV’. A 56-year-old female with a 4-year history of
thrombocytosis. Previous bone marrow trephine samples reported as
‘essential thrombocythaemia-like’ at initial presentation, with lower
cellularity but noted pleomorphism among megakaryocytes. This
sample is hypercellular due to panmyelosis, with marked pleomor-
phism and clustering of prominent megakaryocytes. H&E stain; orig-
inal magnification 920.
Review
340 ª 2015 John Wiley & Sons Ltd British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions (https://onlinelibrary.w
iley.com /term
nline L ibrary for rules of use; O
A articles are governed by the applicable C
reative C om
m ons L
survived at 18 months from first presentation and the main
cause of death was thrombosis (Chievitz & Thiede, 1962).
However, in those who survive longer, transformation to
myelofibrosis and acute leukaemia are life-limiting events
(Kiladjian et al, 2011).
In treating PV, our aim is to reduce complications and
thus improve survival. Thromboembolic events are the major
cause of morbidity and mortality; consequently, their preven-
tion is the main aim of any treatment strategy. A difficulty
in assessing the ability of novel therapies to address this is
the lack of an appropriate surrogate endpoint for the mea-
surement of prevention of thromboembolic complications
and the low prevalence of these complications in patients
treated with current therapies. However we would wish to
emphasize that rates of thromboembolic events remain
higher than the normal population and are an appropriate
target. Prevention of progression to myelofibrosis or acute
leukaemia, and avoiding administration of any treatment that
would increase disease-related tendency to progression would
also be desirable. Fulfilment of these aims is presently an
aspiration, as it would require treatments with demonstrated
ability to prevent progression and no such therapies have as
yet been discovered. Nonetheless, it remains a valuable aspi-
ration when evaluating novel therapies, as these targets are
valid and worthy of perusal.
While the main aim of therapy in PV is directed at the
prevention of thromboembolic events, in the past decade
there has been an increasing appreciation of the additional
individual and societal burdens associated with impaired
quality of life in PV patients, manifest not only by sequelae
of thrombotic events but also by disease-related symptoms.
Large studies have demonstrated a high prevalence of symp-
toms and impaired quality of life in specific patient cohorts.
These studies have also established and validated tools for
the quantitation of such effects (Geyer & Mesa, 2014; Scher-
ber et al, 2014). To what extent these symptoms will respond
to standard therapies and differ from symptoms reported by
case-matched controls remains to be assessed formally and is
important for the field. Nonetheless, it is clear that some
patients have a high burden of symptoms (often pruritus)
that are clearly related to disease and not ameliorated by
standard treatments. These issues have been brought more
into focus by the development of JAK inhibitors, one benefit
of which appears to be to successful control of disease-related
symptoms. This is discussed in more detail below but the rel-
evance of symptom control as an aim of therapy merits its
inclusion as a separate item for consideration here.
Risk groups: what is high-risk PV?
Treatment aimed at the prevention of thromboembolic
events can involve measures such as venesection to a target
Hct of 045 (rarely, lower targets are used in symptomatic
Table II. Criteria for the diagnosis of post-polycythaemia vera
myelofibrosis (PPV-MF) (Barosi et al, 2008a,b).
Major criteria – both required 1 Documentation of a previous diagnosis of polycythaemia vera
as defined by the World Health Organization criteria
2 Bone marrow fibrosis grade 2–3 (on 0–3 scale) or grade 3–4 (on 0–4 scale)
Minor criteria – at least two required 1 Anaemia or sustained loss of requirement of either phlebotomy
(in the absence of cytoreductive therapy) or cytoreductive
treatment for erythrocytosis
3 Increasing splenomegaly, defined as either an increase in
palpable splenomegaly of ≥5 cm (distance of the tip of the
spleen from the left costal margin) or the appearance of a
newly palpable splenomegaly
(>375°C)
A patient with longstanding polycythaemia vera, now thrombocy-
topenic, unresponsive to hydroxycarbamide and anagrelide. Biopsy is
hypercellular with obvious collagen fibrosis. At this stage the histo-
logical picture is indistinguishable from primary myelofibrosis. H&E
stain; original magnification 920. (B) PPV-MF. Gomori reticulin
stain to accompany H&E shown in panel A; original magnification
920.
Review
ª 2015 John Wiley & Sons Ltd 341 British Journal of Haematology, 2016, 172, 337–349
13652141, 2016, 3, D ow
nloaded from https://onlinelibrary.w
iley O nline L
s and C onditions…
Related Documents
