research paper
The effects of hydroxycarbamide on the plasma proteome of
children with sickle cell anaemia
John Brewin,1 Sanjay Tewari,1 Stephan
Menzel,1
Fenella Kirkham,2,3 Baba
4
Inusa, George Renney,5 Malcolm Ward5
and David C. Rees1
1
Red Cell Biology Unit, King’s College
Hospital, King’s College London,
2
Developmental Neurosciences and
Biomedical Research Centre, UCL Great
Ormond Street Institute of Child Health,
London, 3Clinical and Experimental
Sciences, University of Southampton,
Southampton, 4Paediatric Haematology,
Evelina Children’s Hospital,Guy’s and St
Thomas’ Hospital, London and 5Proteomics
Laboratory, Institute of Psychiatry, King’s
College London, UK
Received 21 January 2019; accepted for
publication 10 April 2019
Correspondence: David Rees, Department of
Haematological Medicine, King’s College
Hospital, Denmark Hill, London SE5 9RS, UK.
E-mail: david.rees2@nhs.net
Summary
We investigated changes in the plasma proteome of children with sickle cell
anaemia (SCA) associated with hydroxycarbamide (HC) use, to further
characterize the actions of HC. Fifty-one children with SCA consented to
take part in this study. Eighteen were taking HC at a median dose of
22 mg/kg, and 33 were not on HC. Plasma was analysed using an unbiased
proteomic approach and a panel of 92 neurological biomarkers. HC was
associated with increased haemoglobin (Hb) (898 vs. 814 g/l, P = 0007)
and HbF (67 vs. 153%, P < 0001). Seventeen proteins were decreased on
HC compared to controls by a factor of 13 increased concentration. HC use was associated with reduced haemolysis (lower a, b, d globin chains, haptoglobin-related protein, complement
C9; higher haemopexin), reduced inflammation (lower a-1-acid glycoprotein, CD5 antigen-like protein, ceruloplasmin, factor XII, immunoglobulins,
cysteine-rich secretory protein 3, vitamin D-binding protein) and decreased
activation of coagulation (lower factor XII, carboxypeptidase B2, platelet
basic protein). There was a significant correlation between the increase in
HbF% on HC and haemopexin levels (r = 0603, P = 0023). This study
demonstrated three ways in which HC may be beneficial in SCA, and identified novel proteins that may be useful to monitor therapeutic response.
Keywords: sickle cell anaemia, hydroxycarbamide, proteomics.
Introduction
The most prevalent and severe type of sickle cell disease is
caused by homozygosity for the sickle mutation (HBB;
c.20A>T, p.Glu7Val), and referred to as sickle cell anaemia
(SCA). It is one of the commonest severe inherited disorders
in the world, and is associated with unpredictable acute complications, progressive organ damage and shortened life
expectancy (Brousse et al, 2014). Although outcomes have
improved significantly in high-income countries over the last
40 years, there are still very few treatments available, with
hydroxycarbamide (HC) continuing to be the only drug
which has convincingly been shown to modify the natural
history of the condition and prevent potential organ damage
(McGann & Ware, 2015).
Randomized controlled trials have shown that HC reduces
the frequency of acute pain and acute chest syndrome (Charache et al, 1995; Wang et al, 2011), reduces the need for
blood transfusion (Charache et al, 1995), reduces the risk of
ª 2019 British Society for Haematology and John Wiley & Sons Ltd
British Journal of Haematology, 2019, 186, 879–886
infection and malaria in some settings (Tshilolo et al, 2018),
and is effective as part of the management of abnormal transcranial Doppler (TCD) velocities (Ware et al, 2016). Uncontrolled studies also suggest that HC increases oxygen
saturations (Singh et al, 2008), decreases albuminuria (Tehseen et al, 2017) and may increase life expectancy in adults
(Steinberg et al, 2003). In the USA and UK it is recommended that HC is offered to all children with SCA (Yawn
et al, 2014; Qureshi et al, 2018), and approximately 80%
children take it in some large centres in the USA; in Europe
practice varies widely, but HC is offered to most symptomatic children.
The main mechanism of action of HC is to increase haemoglobin F (HbF) production, which decreases the rate of
HbS polymerization, with subsequent improvement in many
downstream pathologies, including reductions in vaso-occlusion, inflammation, anaemia and haemolysis (Ware, 2010).
HC has some other actions beyond the b-globin locus,
including reducing white cell numbers, decreasing expression
First published online 29 May 2019
doi: 10.1111/bjh.15996
J. Brewin et al
of adhesion molecules on the vascular endothelium and possibly improving nitric oxide metabolism (Rees, 2011),
although it is unclear how much these contribute to its therapeutic actions. In a mouse model of sickle cell disease in
which HC caused no increase in HbF levels, there was no
evidence of improvement in clinical or laboratory parameters
(Lebensburger et al, 2010).
Another important unanswered question involves variability in the therapeutic response to HC. It is unclear why some
patients appear not to respond as well as others. Some of this
may be related to drug adherence, although all studies show
a variable response. In the HUSTLE study, the HbF level varied from 162–278% after one year in patients achieving the
maximum tolerated dose (Estepp et al, 2017); some of this
variability is related to baseline HbF levels, although analysis
of data from the HUG-KIDS study showed that this
explained only about 18% of the HbF response (Ware et al,
2002).
Previous studies have measured the effects of HC on
selected biomarkers, such as indicators of inflammation
(Penkert et al, 2018) or vasculopathy (Lapoumeroulie et al,
2005), but we have adopted an unbiased proteomic
approach to identify changes in plasma protein levels associated with HC use. We hypothesized that there would be
significant differences in the plasma proteome between children with SCD taking and not taking HC, and that these
differences would shed light on the mechanism of action of
HC and may identify novel biomarkers associated with
response to HC.
Methods
Patients and setting
The data were collected as part of a study of silent cerebral
infarcts (SCIs), funded by the Stroke Association (Grant TSA
2012/06) (Tewari et al, 2018). The study was approved by
the UK National Research Ethics Committee (reference 13/
LO/0709) and all patients/parents gave written consent. Children with SCA were recruited from clinics at King’s College
Hospital and the Evelina Children’s Hospital, London. The
aim was to recruit 50 patients, with approximately equal
numbers of those with SCIs and controls with normal brain
magnetic resonance imaging (MRI) scans. Children already
known to have SCIs from previous MRI scans were selectively recruited and eligible patients without previous MRIs
were recruited sequentially as they attended clinic. Children
taking HC were recruited to both arms and had been on a
stable dose of HC for at least 6 months. Inclusion criteria
were: sickle cell anaemia (HbSS), age 8–18 years old, normal
or conditional TCD velocities (13-fold higher
concentrations in children taking HC. It is difficult to know
the significance of increased levels of a-1B-glycoprotein and a2-macroglobulin: a-1B-glycoprotein is an immunoglobulinlike protein of unknown function and a-2-macroglobulin is a
multi-functional proteinase inhibitor, which has been found at
881
J. Brewin et al
Table I. Comparison of laboratory and clinical measurements of those taking hydroxycarbamide with those who were not
Controls
Age (years)
Hb (g/l)
MCV (fl)
HbF (%)
HbF (g/l)
Neutrophil count (3109/l)
Reticulocyte count (3109/l)
Bilirubin
AST
LDH
Hydroxycarbamide
n
Mean
SD
n
Mean
SD
P value
33
33
33
33
33
33
33
33
33
32
116
814
832
67
55
50
389
427
598
601
383
729
743
414
36
150
922
163
245
141
18
18
18
18
18
18
18
18
18
18
124
898
943
153
141
38
227
389
517
505
250
1085
1370
802
818
181
891
194
219
113
034
0007
0004
Purchase answer to see full
attachment