Ann Hematol (2016) 95:1881–1886
DOI 10.1007/s00277-016-2791-y
ORIGINAL ARTICLE
Cyclosporin A for persistent or chronic immune
thrombocytopenia in children
Anthony P. Y. Liu 1 & Daniel K. L. Cheuk 1 & Ana H. Y. Lee 1 & Pamela P. W. Lee 1 &
Alan K. S. Chiang 1 & S. Y. Ha 1 & W. C. Tsoi 2 & Godfrey C. F. Chan 1
Received: 31 May 2016 / Accepted: 5 August 2016 / Published online: 15 August 2016
# Springer-Verlag Berlin Heidelberg 2016
Abstract Twenty percent of children with immune thrombo-
cytopenia (ITP) develop a chronic course where treatment
strategy is less established. Cyclosporin A (CSA) has been
shown to be effective in small series of children with chronic
ITP and might reduce the need for chronic steroid therapy and/
or splenectomy. We reviewed consecutive patients below
18 years old with persistent or chronic ITP treated with CSA
in our unit between January 1998 and June 2015. Thirty
patients (14 boys and 16 girls) were included. The median
age at initial diagnosis of ITP was 5 years (range 0.5–
16.2 years). CSA was started at a median of 13.9 months (range
3.4–124 months) after initial diagnosis and given for a median
duration of 9.3 months (range 0.2–63.9 months). The median
platelet count before commencement was 12 × 109/L (range 4–
199 × 109/L). The median dose of CSAwas 6 mg/kg/day (range
2.4–7.5 mg/kg/day). Complete response (CR) or response (R)
was achieved in 17 patients (57 %), and 7 (23 %) had sustained
response. Side effects (most commonly hirsutism) were tolera-
ble and reversible. CSA appeared effective in about half of
persistent or chronic ITP patients and safe as a second-line
agent in managing these children.
Keywords Cyclosporin . Pediatric . Chronic . Immune
thrombocytopenia
* Anthony P. Y. Liu
apyliu@hku.hk
1
Department of Paediatrics and Adolescent Medicine, Li Ka Shing
Faculty of Medicine, The University of Hong Kong, Queen Mary
Hospital, Hong Kong Special Administrative Region, China
2
Hong Kong Red Cross Blood Transfusion Service, Hong Kong
Special Administrative Region, China
Introduction
Immune thrombocytopenia (ITP) is an autoimmune process
with accelerated destruction of otherwise normal platelets in
response to infectious or other stimulus [1]. Reported annual
incidence in the pediatric population varied between 1.9 and
6.5 per 100,000 children [2]. Although majority of children
have their ITP resolved spontaneously within 3 months, 20 %
have persistent thrombocytopenia beyond 6 months from
diagnosis [3]. In 2008, the International Working Group
reclassified the phases of ITP into Bnewly diagnosed ITP^
(within 3 months from diagnosis), Bpersistent ITP^ (3–
12 months from diagnosis), and Bchronic ITP^ (lasting for
more than 12 months from diagnosis) [4].
While options for treating newly diagnosed ITP are
those of expectant observation, IVIg, anti-D, and cortico-
steroid, the optimal management strategies for patients
with persistent or chronic ITP are less well established.
Second-line treatment options suggested in recent consen-
sus guidelines included splenectomy, rituximab, high-dose
corticosteroid, cyclosporin A (CSA), eltrombopag, azathi-
oprine, vinca alkaloids, and danazol [1, 5]. Chronic or
recurrent steroid use in children is associated with myriad
adverse effects such as growth failure, reduced bone
density, cataract, and immunosuppression; whereas sple-
nectomy, although being effective in 70 % of children
with chronic ITP, carries the immediate risk of surgery
as well as lifelong risk of sepsis and is not recommended in
children younger than 3–5 years old [6]. CSA, a calcineurin
inhibitor widely applied to patients with autoimmune condi-
tions and in the post-transplant setting, has been trialed in
pediatric chronic ITP with some success but evidence is still
scarce [7–10]. We hereby report our experience in CSA use
for children with persistent or chronic ITP, which represents
the largest pediatric series to date.
1882
Method
We retrospectively reviewed the patient database of the
Children’s Centre for Cancer and Blood Diseases,
Department of Paediatrics and Adolescent Medicine, Queen
Mary Hospital in Hong Kong. All patients with persistent or
chronic ITP (disease lasting for more than 3 months from
diagnosis), diagnosed at <18 years of age between 1 January
1998 and 30 June 2015, and treated with CSA were included.
CSA was indicated as second-line treatment for profound or
symptomatic thrombocytopenia without adequate response to
IVIg and steroid, as steroid-sparing agent for patients with
steroid dependence, or in patients with persistent or chronic
ITP who elected not to receive steroid or IVIg.
Information on patient demographics, presenting symp-
toms and platelet count, investigations (platelet autoanti-
bodies, bone marrow examination, anti-nuclear antibodies
[ANA]), treatment prior to CSA use, timing and dosage of
CSA, treatment response to CSA, adverse effects of the drug,
and requirement of further alternative treatment strategies
were retrieved and documented from the patient files and
hospital electronic patient record system. A qualitative solid
phase enzyme linked immunosorbent assay (LIFECODES
PakAuto, Immucor GTI Diagnostics Inc., Waukesha, WI,
USA) was used to detect the presence of platelet glycoprotein-
specific autoantibodies in patients’ serum and autoantibodies
eluted from their platelets. CSA was started at a dose of
3–6 mg/kg/day and titrated to the target drug level of
100–200 μg/L, platelet count, and adverse effects.
Data analysis was primarily descriptive. Continuous vari-
ables were expressed as median and range. In accordance with
the IWG recommendation, complete response (CR) was de-
fined as platelet count ≥100 × 109/L and absence of bleeding,
and response (R) as platelet count ≥30–<100 × 109/L and at
least doubling the baseline count as well as absence of bleeding.
No response (NR) as platelet count <30 × 109/L or less than
twofold increase from the baseline count or presence of bleed-
ing despite 3 months of CSA treatment [4]. Patients in which
CSA had to be terminated due to adverse effect(s) before
response could be observed were incorporated into the NR
group. Patients who achieve CR are considered to have
sustained response if CR is maintained during and after treat-
ment with CSA. The Fisher’s exact test and the Wilcoxon rank-
sum test were used to compare characteristics of CSA re-
sponders and non-responders; p values <0.05 were deemed
statistically significant.
Results
Thirty children with persistent or chronic ITP were treated
with CSA during the study period. Fourteen (47 %) were
boys, and the median age at initial diagnosis of ITP was
Ann Hematol (2016) 95:1881–1886
5.0 years (range 0.5–16.2 years). The median initial platelet
count was 9 × 109/L (range 2–58 × 109/L). Marrow examination
was performed in 29 patients with all of them showing mega-
karyocytic hyperplasia and absence of abnormal infiltration sug-
gestive of peripheral consumption. Platelet glycoprotein-specific
autoantibodies were detected in the patients’ serum and/or plate-
let eluates for all the 21 patients in whom the test was performed.
ANA were positive in 11 of 27 patients tested. Concomitant
autoimmune hemolytic anemia was present in two patients.
None of the patients had identifiable secondary causes, and all
the cases were therefore considered as Bidiopathic^ in this series.
During the disease course, two patients experienced life-
threatening bleeding (one with intra-cranial hemorrhage and
one with gastrointestinal bleeding).
The treatment used prior to initiation of CSA is summa-
rized in Fig. 1. Among the 28 patients who received prednis-
olone, CR was achieved in 18, R in 4, and NR in 6. CSA was
started for steroid dependence in 18 patients, refractoriness or
suboptimal response to steroid in 10 patients, and at the
treating clinicians’ preference in 2 patients who had not been
treated with steroid. The drug was started at a median of
13.9 months (range 3.4–124 months) after the initial diagnosis
of ITP and given for a median of 9.3 months (range 0.2–
63.9 months). The median platelet count before commence-
ment of CSA was 12 × 109/L (range 4–199 × 109/L). The me-
dian dose of CSA was 6 mg/kg/day (range 2.4–7.5 mg/kg/day),
and the target drug level of 100–200 μg/L was met in 12 out of
25 patients (48 %) where the information was available. The
cohort was followed-up for a median duration of 6.5 years
(range 0.1–12.1 years) after the initiation of CSA.
After treatment with CSA, CR was achieved in 11 (37 %)
patients (one with steroid), R was achieved in 6 (20 %)
patients (one with steroid). Thirteen (43 %) patients had NR.
The median duration between the start of CSA and CR or best
platelet response in patients with R, was 2.8 months (range
0.3–9.6 months). Out of the 17 patients with CR or R, 7 had
sustained response and remained in remission after treatment
with CSA alone (1 patient required 2 courses of CSA).
Among these seven sustained responders, the median duration
of CSA use was 23.9 months (range 18.9–59 months); and
after excluding a patient who was lost to follow-up upon stop-
ping of CSA, the median duration of remission after cessation
of the drug was 41.7 months (range 5.5–116.4 months). Age,
sex, ANA positivity, platelet autoantibody testing, duration of
disease at initiation of CSA, initial platelet count, platelet
count at start of CSA, and previous response to steroid as well
as IVIg did not predict treatment outcome with CSA.
Side effects of CSA were in general tolerable and reversible,
namely hirsutism (n = 18), tremor (n = 3), hypertension (n = 2),
impaired renal function (n = 2), gum hypertrophy (n = 1), and
neutropenia (n = 1). CSA was terminated in one patient due to
reversible elevation in creatinine (20μmol/L from baseline)
despite a relatively low CSA dosage (3 mg/kg/day) given for
Ann Hematol (2016) 95:1881–1886 1883

Fig. 1 Treatment response to TREATMENT PRIOR TO CSA

CSA in the study cohort
No prior IVIg only Steroid only Steroid + IVIg Steroid + IVIg +
treatment +/- an-D other agents
N=1 N=1 N=2 N = 22 N =4

Children with
Persistent/Chronic ITP
receiving CSA
N = 30

CR R NR
N = 11 N=6 N = 13

Sustained Relapsed Sponta-

response requiring Further Tx Further Tx neous
Further Tx recovery
N=7 N=4 N=6 N = 10 N =3

CR R CR R CR R NR*
N =3 N=1 N =5 N=1 N=6 N=2 N=2

* No response despite use of mulple treatment strategies

and remained thrombocytopenic (< 30 x 109/L)

1 week only. Other than those with sustained CR to CSA,
further treatment was given in 20 patients. At the end of the
study period, 23 patients out of the entire cohort were in CR,
with 5 in R, and 2 in NR. The final treatment that led to CR/R
included CSA (n = 7), CSA with steroid and eltrombopag
(n = 1), splenectomy (n = 9), spontaneous remission (n = 6),
rituximab (n = 2), eltrombopag (n = 2), and MMF (n = 1).
Discussion
Since the initial reports of CSA use for ITP in the 1980s,
limited data has been published with regard to its efficacy in
the condition. A randomized controlled trial comparing com-
bination therapy with CSA and thrombopoietin agonist versus
thrombopoietin agonist alone in adult patients with refractory
ITP showed similar response rate but significantly lower
relapse rate in the combination arm [11]. Controlled trial for
pediatric age group is however lacking, with case series
reporting response rates from 18 to 80 % using CSA as mono-
therapy or combination therapy (Table 1) [7, 10, 12–15]. Our
data represented the largest consecutive cohort published. The
response rate of 57 % and sustained response rate of 23 % in
the current study were in line with the results by Perrotta and
colleagues, who treated 14 patients with CSA at a median
dosage of 10 mg/kg/day, resulting in a response rate of 50 %
and sustained response in 28.6 % lasting between 38 and
62 months [10]. On the contrary, response rate was only 27 %
in the 15 pediatric ITP patients treated by Choudhary DR and
colleagues with CSA at 5 mg/kg/day for 1 week, followed by
dosage at 3 mg/kg/day [12]. The above suggested that CSA
might have more effect in children with persistent or chronic
ITP at a higher dosage (6–10 mg/kg/day). Combination treat-
ment with CSA (10 mg/kg/day, for median of 4 months), vin-
cristine (1.5 mg/m2 weekly, for 2–4 doses), and methylprednis-
olone (100 mg/m2 weekly, for 2–4 doses) was adopted in 10
pediatric patients with a promising complete response rate of
70 % and median time to platelet count greater than 50 × 109/L
of 7 days. These patients were followed for a median of
13 months (9–37) after completion of treatment. In accordance
with our findings, no demographic or clinical predictors of
response to CSA use were identified in the literature of pediatric
ITP (including age at diagnosis, age at CSA use, pre-treatment
platelet levels, duration of disease, number and modalities of
prior treatment) [10, 12]. Adverse effects observed were
generally reversible with hypertrichosis being the most
common one, although opportunistic infection had rarely
been reported [10, 12]. The data supported the safe appli-
cation of CSA as a second line agent for children with
chronic ITP, sparing them from the adverse effects of
long-term steroid exposure and/or splenectomy.
Rituximab was recommended as a second line treatment
option in both the ASH and the ICR guidelines, but random-
ized controlled trial on its efficacy versus placebo or CSA was
unavailable [1, 5]. A systematic review summarized 14 studies
on pediatric ITP (n = 323) showing a pooled response rate
Table 1 Case series on cyclosporin use in pediatric chronic ITP
1884

Subject M/F Median Inclusion Duration Median Steroid CSA dosage Target Overall CR PR Assessment of S/E
number age of ITP Plt at response CSA level response response
(years) (m) initiation (ug/L) (CR + R) durability
(range) (range) of CSA
(×109/L)
(range)

Choudhary Total: 25 15:10 12 Primary chronic ITP 24 18 4/25 5 mg/kg/day 200–400 27 % 20 % 7 % (1/15) in Mean FU in Gingival hyperplasia
DR et al. Pediatrics: (4–73) Platelet <30 × 109/L (6–90) (2–42) for 1 week, (4/15) in (3/15) in <18 years those with in 2 patients, out of
[12] 15 or bleeding then 3 mg/kg/ <18 years <18 years response 7.3 m entire cohort
day (±2.1 m)
Williams 10 5:5 16 Primary refractory/ 13.5 7 0/10 10 mg/kg/day 100–200 80 % 70 % (Plt 10 % Mean FU 13 m Nausea, hirsutism,
JA et al. (10–18) chronic ITP (1–35) (1–24) in normal (Platelet 80– (9–37 m) after peripheral
[13] Evans syndrome range after 120 × 109/L completion of neuropathy,
(4/10)a stopping for ≥3 m CSA constipation, jaw
treatment) after stopping aches, alopecia
treatment)
Perrotta S 14 7:7 8.5 Chronic ITP 24 11.5 N/A Median 100–250 50 % 29 % 21 % (Rise 38–62 m in Hypertrichosis in 4,
et al. [10] (5–17) Platelet <20 × 109/L (6–90) (4–19) 10 mg/kg/day (Rise of of Plt 50– patients with transient
9
(7.5–10) Plt >150 × 150 × 10 /L) CR; 38–62 m; hyperbilirubineemia
109/L) 1 m in patients in 1, systemic
with PR mycosis in 1
Schultz KR 5 4:1 11 ITP dependent on 3 N/A N/A Started at >150 At 5 mg/kg/day: 1 had increase of Plt 1–4 weeks, Plt Headache,
et al. [8] (3–14) immunosuppressants (3–10) 5 mg/kg/day ≥25 × 109/L declined in all 4 gastrointestinal
for ≥3 m for 2 weeks above baseline with response upset, decreased
and increased At 10 mg/kg/day: 1 had increase of Plt while on CSA appetite, irritability,
to 10 mg/kg/ ≥25 × 109/L chest pain,
day if no above baseline; 2 had increase of Plt peripheral neuritis,
response ≥50 × 109/L hypertension
above baseline
Current 31 15:16 5.1 Primary persistent/ 12 12 23/29 6 mg/kg/day 100–250 55 % 35 % (Plt 19 % (Plt 1 lost to FU Hirsutism (n = 18),
study (0.5– chronic ITP (2.5–124) (2–199) (2.4–7.5) ≥100 × ≥30–< 100 × median FU tremor (n = 3),
16.2) 109/L and 109/L and at 41.7 m (5.5– hypertension (n = 2),
no least 116.4) in the impaired renal
bleeding) doubling the remaining 6 function (n = 2),
baseline with sustained gum hypertrophy
count, and response after (n = 1) and
absence of completion of neutropenia (n = 1)
bleeding) CSA

a
Combination treatment with CSA, VCR 1.5 mg/m2 weekly, MP 100 mg/m2 weekly, till Plt >50 (2–4 doses)
Ann Hematol (2016) 95:1881–1886
Ann Hematol (2016) 95:1881–1886
(platelet ≥30 × 109/L) and complete response rate (plate-
let ≥100 × 109/L) of 68 and 39 %, respectively, with the
combined median time to response being 3 weeks and
median response duration being 12.8 months [16]. Possible
predictors of response to rituximab were shorter duration of
disease, steroid responsiveness, and secondary ITP. Adverse
reactions such as allergy and serum sickness occurred in
41.1 %. Notably, patients had developed immunodeficient
states and significant infections after rituximab use, including
varicella, pneumonia, and enteroviral meningoencephalitis. In
addition, reactivation of hepatitis B virus upon rituximab use
was well described in individuals with chronic carrier state
which in turn was highly prevalent in Southeast Asia [17].
More recently, data supporting the use of thrombopoietin
receptor agonist, namely eltrombopag, in pediatric ITP had
become available leading to the FDA approval in 2015 for
its use in children with chronic ITP as young as 1 year of
age refractory to steroid, immunoglobulin or splenectomy
[18, 19]. PETIT was a phase II trial which enrolled 52 children
to be randomized to 7 weeks of eltrombopag or placebo
followed by open label eltrombopag for a total of 24 weeks.
Significantly, higher proportion of patients randomized to re-
ceive eltrombopag achieved platelet count >50 × 109/L com-
pared with placebo (63 vs 18 %, respectively; p = 0.0043).
Safety profile for eltrombopag was favorable but one should
remain vigilant against possible liver dysfunction and cataract.
In the phase III PETIT2 trial, 92 children with chronic ITP
were first randomized to receive 13 weeks of eltrombopag or
placebo followed by open label eltrombopag for total of
24 weeks, significantly more patients achieved consistent
response when compared with placebo (43 vs 4 %, respectively;
p = 0.0011). Use of eltrombopag also decreased the need for
rescue treatment and allowed reduction of baseline ITP medica-
tions, but further follow-up data would be required to determine
the durability of treatment response after cessation of treatment.
On the practical side, financial implications for prolonged use of
eltrombopag would be substantial for families with children
with chronic ITP. For an average-sized 6-year-old Chinese boy
(body surface area ∼0.8 m2) managed in our unit, the cost of
eltrombopag at 25 mg daily for one year would be approximate-
ly US$78,000, while the cost for treating the same patient with
either a course of rituximab (375 mg/m2 for 4 doses) or 1 year of
CSA (6 mg/kg/day) would be US$6650 and US$3650,
respectively.
Our current study represented the largest cohort in the eval-
uation of treatment response of pediatric ITP to CSA, and
there were several strengths to it. We included only patients
with primary chronic ITP to minimize subject heterogeneity,
and two thirds of patients had their diagnosis supported by
platelet autoantibodies assay. Disease response was evaluated
in accordance with the latest IWG recommendations allowing
more consistent comparison. On the other hand, there were a
number of limitations to this review. Despite consecutive
1885
enrollment of patients, the single institution experience might
incur referral bias. Our retrospective, single-armed study also
did not allow comparison of efficacy between CSA and other
agents or placebo. In addition, treatment was individualized in
our cohort, and there was no standardized timing for initiation
and titration and duration of use of CSA. Further studies on the
efficacy and cost-effectiveness of CSA in comparison with other
agents in pediatric ITP in the form of a randomized-controlled
trial would be useful in determining the place of CSA and other
second-line agents in management of pediatric ITP.
Conclusion
CSA is a relatively inexpensive agent in treating children with
persistent or chronic ITP and appears safe and effective in a
substantial proportion of patients. Response to the agent
would obviate the adverse effects from chronic steroid use
and splenectomy in these young patients. Further research in
establishing its cost-effectiveness and efficacy in comparison
with other second-line agents is needed.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
References
1. Neunert C, Lim W, Crowther M et al (2011) The American Society
of Hematology 2011 evidence-based practice guideline for immune
thrombocytopenia. Blood 117:4190–4207
2. Terrell DR, Beebe LA, Vesely SK et al (2010) The incidence of
immune thrombocytopenic purpura in children and adults: a critical
review of published reports. Am J Hematol 85:174–180
3. Kühne T, Buchanan GR, Zimmerman S et al (2003) A prospective
comparative study of 2540 infants and children with newly diag-
nosed idiopathic thrombocytopenic purpura (ITP) from the
Intercontinental Childhood ITP Study Group. J Pediatr 143:605–608
4. Rodeghiero F, Stasi R, Gernsheimer T et al (2009) Standardization
of terminology, definitions and outcome criteria in immune throm-
bocytopenic purpura of adults and children: report from an interna-
tional working group. Blood 113:2386–2393
5. Provan D, Stasi R, Newland AC et al (2010) International consen-
sus report on the investigation and management of primary immune
thrombocytopenia. Blood 115:168–186
6. Aronis S, Platokouki H, Avgeri M et al (2004) Retrospective eval-
uation of long-term efficacy and safety of splenectomy in chronic
idiopathic thrombocytopenic purpura in children. Acta Paediatr 93:
638–642
7. Gesundheit B, Cividalli G, Freeman A et al (2001) Cyclosporin A
in the treatment of refractory immune thrombocytopenia purpura in
children. Eur J Haematol 66:347–351
8. Schultz K, Strahlendorf C, Warrier I, Ravindranath Y (1995)
Cyclosporin A therapy of immune-mediated thrombocytopenia
in children
1886
9. Jain S, Kapoor S, Singh A et al (2012) Good response to cyclosporin
in a child with acquired thrombocytopenic purpura. Indian J Pediatr
79:1236–1237
10. Perrotta S, Amendola G, Locatelli F et al (2003) Treatment with short-
term, high-dose cyclosporin A in children with refractory chronic
idiopathic thrombocytopenic purpura. Br J Haematol 121:143–147
11. Cui ZG, Liu XG, Qin P et al (2013) Recombinant human
thrombopoietin in combination with cyclosporin A as a novel ther-
apy in corticosteroid-resistant primary immune thrombocytopenia.
Chin Med J (Engl) 126:4145–4148
12. Choudhary DR, Naithani R, Mahapatra M et al. Efficacy of cyclo-
sporine as a single agent therapy in chronic idiopathic thrombocy-
topenic purpura. Haematologica 2008; 93: e61-62; discussion e63
13. Williams JA, Boxer LA (2003) Combination therapy for refractory
idiopathic thrombocytopenic purpura in adolescents. J Pediatr
Hematol Oncol 25:232–235
14. Moskowitz IP, Gaynon PS, Shahidi NT, Cripe TP (1999) Low-dose
cyclosporin A therapy in children with refractory immune throm-
bocytopenic purpura. J Pediatr Hematol Oncol 21:77–79
Ann Hematol (2016) 95:1881–1886
15. Schultz KR, Strahlendorf C, Warrier I, Ravindranath Y (1995)
Cyclosporin A therapy of immune-mediated thrombocytopenia in
children. Blood 85:1406–1408
16. Liang Y, Zhang L, Gao J et al (2012) Rituximab for children with
immune thrombocytopenia: a systematic review. PLoS One 7:e36698
17. Pei S-N, Chen C-H, Lee C-M et al (2010) Reactivation of hepatitis
B virus following rituximab-based regimens: a serious complica-
tion in both HBsAg-positive and HBsAg-negative patients. Ann
Hematol 89:255–262
18. Bussel JB, de Miguel PG, Despotovic JM et al (2015) Eltrombopag
for the treatment of children with persistent and chronic immune
thrombocytopenia (PETIT): a randomised, multicentre, placebo-
controlled study. Lancet Haematol 2:e315–e325
19. Grainger JD, Locatelli F, Chotsampancharoen T et al. Eltrombopag
for children with chronic immune thrombocytopenia (PETIT2): a
randomised, multicentre, placebo-controlled trial. The Lancet 2015