6 J. E. Vaughn et al.
was ineffective as well. Administration of a thrombopoi-
etin (TPO) receptor agonist (eltrombopag) resulted in rising
platelet counts, but treatment had to be discontinued
because of severe neuropathy and myalgia. A second
attempt at lower doses failed to improve counts.
The patient subsequently suffered an intracranial
haemorrhage. She underwent 12 days of plasmapheresis
with no improvement of her thrombocytopenia and was
started on a combination of azathioprine, mycophenolate
mofetil (MMF) and cyclosporine [3], to which weekly
rituximab was added. At the time of our consultation,
there was no response, her platelet count remaining
below 5 9 109/l, and there was evidence of alloimmu-
nization to platelets.
Given the refractory nature of her disease and the
recent intracranial haemorrhage, the question was raised
as to whether the patient would benefit from HCT. We
will review the literature on HCT for ITP and Evans syn-
drome, which comprises ITP-like thrombocytopenia com-
bined with haemolytic anaemia.
Treatment options
Conventional therapy
The majority of patients with ITP will achieve remissions
with the use of steroid therapy, but relapse is frequent [4].
Second-line strategies include IVIG infusion, the anti-CD20
antibody, rituximab and surgical splenectomy. Modifica-
tions to the choice, dosing, frequency and timing of these
second-line modalities are currently under investigation [5,
6]. Concerns exist about the risk of haemorrhage and infec-
tion in patients who undergo splenectomy or are exposed
to multiple doses of rituximab immunotherapy. A recent
comparison of long-term outcome among chronic ITP
patients after splenectomy as compared to rituximab ther-
apy overall showed no difference between the two modali-
ties regarding mortality or need for hospitalization because
of haemorrhage or treatment-related infections. However,
the risks were increased in patients with previous mucosal
bleeding or multiple medical comorbidities [7]. While mul-
tiple scoring systems, which consider both clinical [8–10]
and laboratory [11, 12] parameters have been developed to
assess bleeding risk in ITP patients, with the objective of
identifying those who should receive aggressive treatment,
none have been validated in a large, heterogeneous patient
population [13].
Approximately 10% of patients will become refractory
to second-line therapies, including immunosuppressive or
cytotoxic agents [14, 15]. Recent advances in the treatment
of refractory ITP include the approval of TPO receptor ago-
nists such as romiplostim [16] and eltrombopag [17, 18],
which yield promising results but require continuous
administration to maintain platelet counts. Side effects
include myalgias, hepatotoxicity, thrombosis and the risk
of haemorrhage after cessation of treatment. Side effects
did lead to discontinuation of therapy in the patient pre-
sented above. There has also been concern about the
development of reticulin fibrosis of the marrow with long-
term exposure [19, 20], although the incidence appears to
be low, and changes are reversible.
The growing number of available therapeutic options
has greatly reduced the number of severely refractory
patients. However, in those patients, futile attempts at
treatment, intolerance to medications and alloimmuniza-
tion to platelet transfusions can prove frustrating for the
clinician and devastating for the patient. As expected,
patients with refractory ITP have a higher mortality rate
than patients who respond to treatment [21, 22].
Rationale for transplantation
HCT as treatment for autoimmune cytopenias, including
ITP, was first conceived following the observation of res-
olution of concurrent autoimmune diseases in patients
who underwent HCT for malignant disorders [23, 24],
findings later substantiated by studies in animal models
[25–29]. Those experiments indicated that immunoabla-
tive therapy led to a reduction in circulating autoreactive
T- and B-cell counts. As a result, peripheral autoreactivity
was halted, and the immune systems of these animals
were essentially ‘reset’. This resetting has been hypothe-
sized to occur through various mechanisms, including
restored thymic elimination of self-reactive cells and
selection of self-tolerant cells [30], and restoration of a
regulatory T-cell population that promotes self-tolerance
[31]. As typically immunoablative strategies are used in
preparation for transplantation, such mechanisms would
also explain the efficacy of autologous stem cells in treat-
ing these patients.
The use of allogeneic donor cells for HCT, of course, is
intended to replace completely the autoreactive immune
system of the patient with a healthy donor-derived
immune system and is intended to provide sustained dis-
ease control related to a ‘graft-versus-autoimmunity’
(GVA) effect [32, 33]. Recent studies suggest that reduced
intensity and so-called nonmyeloablative regimens used
in preparation for allogeneic HCT will avoid the toxicity
associated with high-dose therapy and may still achieve a
GVA effect, as long as at least a mixed chimerism of
donor and host cells is achieved [34]. A particular con-
cern in patients with autoimmune diseases undergoing
HCT is, of course, the susceptibility to infection owing to
the often prolonged immunosuppressive treatment before
coming to HCT. Concern has also been expressed in
regards to the development of secondary autoimmune
© 2015 International Society of Blood Transfusion
Vox Sanguinis (2016) 110, 5–11