

Autologous Stem Cell transplantation
and Rheumatic Diseases Introduction All over the world, Rheumatic Diseases
(RD) are today a problem for the social security systems. The cost of their
management is enormous due to the chronicity of the diseases. Indeed, the
quality of life is poor, the loss of the job very high, and the long term
outcome is poor, despite initial positive responses to immunomodulatory therapies.
( ref ) For this reason, new rational therapies are particularly warranted
in order to stop the disease evolution and avoid the deterioration of the
joints and internal organs. Autologous stem cell transplantation (ASCT)following
high dose ablative chemotherapy is today a reality that has become true with
an international study program. The dosage of drugs, such as cyclophosphamide
(Cy), is significantly limited due to side effects and in particular to bone
marrow toxicity.
For this reason, an international project was
developed with the concept of exceeding that limit then rescuing the patient
with an autologous ASCT. This study was developed along consensus study guidelines
such that currently around 600 patients have received an ASCT mostly within
the context of phase I / II studies.(ref Gerber, Gratwohl, Passegg Tyndall
Best Clin Haematol 2002). Around 450 of these cases are registered in the
combined European Group for Blood and Marrow Transplantation ( EBMT) and European
League Against Rheumatism ( EULAR) data base. This report summarises the results
of this project. It should not be forgotten that long term disease free survival
in diseases such as leukaemia are today possible for many patients because
of carefully performed randomised controlled trials, rather than by a full
understanding of the pathophysiology . It is hoped that ASCT for RD , an intrinsically
toxic and dangerous form of treatment, will become obsolete when new targeted
strategies will be developed in the future.
Bone
Marrow Transplantation and ASCT
One of the first published reports of attempted bone marrow transplantation
(BMT) was in 1939 ( ref) in a young woman with suspected gold salt induced
aplastic anaemia, the gold being given to treat pulmonary tuberculosis. Interestingly
, one of the first reports of BMT coincidentally inducing long term remission
of an AD was in a patient with aplastic anaemia due to gold treatment of rheumatoid
arthritis.(ref). In the past decade , most transplants have been performed
with peripherally harvested hematopoietic stem cells ( HSCs) , with fewer
bone marrow and recently some few cord blood HCS harvesting ( table 1 – EBMT
mega file).From the aspect of treating AD, there are some important differences
between peripheral and bone marrow HSC source . Bone marrow contains fewer
T cells , but requires a general anaesthetic and a longer reconstitution time
due to fewer early progenitors. In small children ( juvenile idiopathic arthritis)
, it is sometimes not possible to mobilise enough stem cells for a peripheral
harvest: it is known that at least 2 X 10 6 CD34 positive cells ( stem and
early progenitors) per kg body weight are needed for reliable hematological
reconstitution. On the other hand a peripheral HSC source requires mobilisation
of stem cells from the marrow using high dose Cy and /or growth factors ,
mostly G-CSF. For AD patients this raises some new aspects. In 5 patients
first attempts to mobilise stem cells were unsuccessful, probably due to previous
marrow toxic treatments . This was not as common as initially anticipated
(ref) and was overcome by combining the graft product with a second mobilisation
, or moving to a bone marrow harvest. Also, G- CSF is a potential trigger
for a flare of AD , and this was observed in some patients , especially RA
( 8/72) (ref). Most of these flares were relatively easy to control, but in
at least one ( multiple sclerosis) a fatal outcome ensued(ref).In some cases
the Cy in doses from 2-4 gms /m2 given as mobilisation resulted in a rapid
improvement of the AD, such that this was adopted as the pre-randomisation
step in phase III studies for RA and SLE ( see below). Much discussion has
taken place concerning the need for intense purging of the graft product of
autoaggressive lymphocytes , as well as the concept that only an allograft
could replace a defective immune system or “cure” an AD through a theoretical
graft versus autoimmune cell reaction. Five years and 600 patients later it
is reasonable to say that the more aggressive regimens such as radiation or
busulphan/cyclophosphamide based were associated with more toxicity , but
with only a trend toward more durable remission (ref). Concerning purging,
again an assocation between severity of purging and infection risk was seen
, without clear advantage. It is known that complete cellular purge either
in vivo or in vitro is not possible with chemotherapy or radiation, so that
at the most an “autoimmune debulking” occurs with autologous ASCT. Some groups
did not manipulate the graft product , and in one study (ref) RA patients
were randomised to either CD34 selection or none, and no advantage was observed
in terms of relapse (ref) ; in fact a trend toward fewer relapses was seen
in the non purged group (ref) , raising the issue of modulating T –cells (
previously called suppressor cells) being needed to control autoreactive processes.
Overall a concept of resetting the immune auto reaction rather than ablating
and inducing tolerance , as was suggested by animal model experiments , has
evolved with this project.
Coincidental
Autoimmune Disease and ASCT
Since the first reports of improvement and even long term remission of RA
following allo ASCT (refs), many retrospective and some prospective case reports
and series have been published ( refs). The original allo ASCT case reports
were followed by similar short and long term benefits following auto ASCT
(table 2). While of high interest, such reports have the disadvantage of publication
bias and incomplete details concerning the AD status and outcome measurement.
More recently negative or non response cases have been published (ref) , and
it is well established that AD may be passively transferred through allo ASCT.
However from these case reports some important observations have occurred.
One woman with RA who received an allo ASCT from her healthy brother to treat
gold induced aplastic anaemia enjoyed a 4 year clinical and serological remission
of the RA , followed by a full relapse. All studied immune competent cells
from the blood and joint were of donor type ( Y chromosome containing), illustrating
the complexity of AD pathophysiology. In another case a man with leukaemia
received an allo ASCT from his HLA identical brother with known but inactive
SLE (ref) . The donor had antibodies to C1q, which later developed but then
disappeared in the recipient post transplant, with no signs of renal or other
manifestations of SLE. Both these cases illustrate that the host environment
/modulatory cell network is as critical in the final expression of AD as the
inciting clones of autoreactive cells.
Joint
& Connective Tissue Diseases and ASCT
The EBMT and the EULAR combined stem cell project in autoimmune disease issued
consensus guidelines (ref) before the first patient was treated and a collaboration
with North America and the International Bone Marrow Transplantation Registry
(IBMTR) quickly forged. Similar data sets on transplanted patients are collected
world wide, the Americas in the IBMTR and the rest of the world in Basel in
the EBMT/EULAR data base. The EBMT/EULAR data base contains data on 450 registrations,
the most commonly transplanted diseases being systemic sclerosis (SSc) , RA
, JIA and SLE (table 3). The data come from 82 transplant centres in 21 countries
and the initial transplant related mortality (TRM) was 7% (ref). This was
higher than the predicted 3% for autoASCT overall and reflects the general
overall unwellness and multiorgan involvement of many RD patients compared
with, say ,breast cancer patients undergoing high dose chemotherapy and ASCT.
In fact a there was a marked difference between AD groups with a TRM of 12.5%
in SSc and only one patient with RA. Also different were response rates and
types. In RA, JIA and SLE more patients responded early but later relapsed
than for SSc.
Systemic
sclerosis
In the first 45 patients , an improvement of 25% or more was seen in 70% of
the patients, with a TRM of 17%. (15). Several protocols were used, mostly
either Cy based ( 4g /m2 Cy mobilisation and Cy 200 mg/kg body weight conditioning
or radiation 8 Gy/Cy 120mg/kg body weight .With further patient recruitment,
the TRM fell to 8.6% , considered to be related to more careful patient selection.Lung
function tended to stabilise and some factors were identified as potentially
hazardous for ASCT eg pulmonary hypertension > 50mmHg mean pulmonary arterial
pressure, severe cardiac involvement, severe pulmonary fibrosis and uncontrolled
systemic hypertension. When such patients were excluded from the analysis,
the TRM was 7% , suggesting that the proposed randomised controlled trial
would be ethical , given a near 50% 5 year mortality of this subgroup of patients.
In the new designed trial, ASTIS (Autologous Stemcell Transplantation International
Scleroderma ) Trial, only diffuse skin SSc patients are enrolled with less
than 4 years of diffuse skin involvement and evidence of progressive and organ
or life threatening disease. The primary end point of this trial is event
free survival at 2 years. The treatment arm is mobilisation with Cy 4g/m2
and G-CSF , followed by CY 200mg/kg body weight conditioning plus ATG and
a CD 34 selected graft. The control arm is monthly IVI pulse CY 750mg/m2 for
12 months.( fig 1) Each arm will have 100 patients. ASTIS is running, and
so far using these selction criteria there have been no transplant related
deaths. Further details are available on the website: www.astistrial.com
, or E-mail astistrial-fps@unibas.ch .
Rheumatoid
Arthritis
Of the 63 patients transplanted, an analysis of the first 72 showed significant
improvement , with 78% achieving an ACR 50 response. This is a composite score
clinical and laboratory parameters which should improve by at least 50%. Most
of the patients had failed at least 3 conventional disease modifying antirheumatic
drugs ( DMARDs ) such as methotrexate or sulphasalazine before the transplant.
Some degree of relapse was seen in 73% of patients post transplant , but was
relatively easy to control in most with drugs which had proven ineffective
pretransplant. The median follow-up was 18(6-40) months , and the majority
of patients received a conditioning regimen of Cy 200mg/m2 alone and received
peripherally harvested stem cells after either G-CSF or Cy/ G-CSF ( equal
numbers ) mobilisation. These phase I and II experiences have been integrated
into a phase III randomised study, the ASTIRA ( Autologous Stemcell Transplantation
International Rheumatoid Arthritis ) Trial. In this trial, will be enrolled
only active RA patients who have failed at least 4 DMARDS (including methotrexate
and anti TNF alpha) with a disease duration between 2-15 years. After stem
celll mobilisation is achieved with Cy 4g/m2 and G-CSF, randomisation occur
to either continued conventional therapy with either methotrexate or leflunamide
or conditioning with Cy 200 mg/m2 and ATG. The graft is not be manipulated
, and maintenance with methotrextae or leflunamide is given. The primary end
point is the number of patients reaching a good or moderate EULAR response
or and ACR 20 at six months. 16 patients in each arm are required, calculated
on a > 50% difference in the two groups. Further details from www.EBMT.org
or E-mail: 114077.1034@compuserve.com
Juvenile
Idiopathic Arthritis
A total of 51 children with idiopathic juvenile arthritis, mostly the systemic
form called Stills disease, have been registered . Most of these cases were
treated in two Dutch centers using a bone marrow obtained stem cell source
and a conditioning protocol of Cy 200mg/kg body weight, TBI 4Gy and ATG (
N Wulffraat , personal communication). In the whole group there were 15 complete
remissions and 3 partial remissions reported. In those attaining remission,
the corticosteroiud dose could be reduced and some patients experienced puberty
and catch up growth. Three patients died from macrophage activation syndrome
, thought to be related to intercurrent infection or uncontrolled systemic
activity of the disease at the time of transplantation. Protocols were modified
accordingly such that systemic activity is controlled before the transplant
with methyl prednisolone intravenously . Since this modification , no further
such deaths have occurred. Further phase I and II pilot studies will be undertaken
before the optimal phase III randomised study will be proposed.
Systemic
Lupus Erythematosus
Of the 55registrations , most had either renal and /or CNS involvement, and
21 had failed conventional Cy treatment. A peripheral stem cell source after
mobilisation with Cy and G-CSF was used in the majority. Twenty -three patients
received a conditioning with Cy and ATG , eleven with Cy and TBI and five
other regimens were employed. There were 5 deaths due to treatment and one
from progressive disease, resulting in an actuarially adjusted TRM of 10(2-20)%.
In those patients with sufficient data for analysis , 72% achieved a “ remission”
, defined as a SLEDAI ( Systemic Lupus Erythematosus Disease Activity Index)
of = 3 and steroid reduction to < 10mg/day. Six others did not reach this
end point. Half of the remissions relapsed to some degree, and were mostly
easily controlled on standard agents which had previously been ineffective.
A more complete analysis of the data is underway with a view to proposing
the most appropriate phase III protocol. Traynor and colleagues (34) reported
on 9 patients with severe SLE, who were mobilised in a transplant protocol.
One died as a result of infection following mobilisation and another 3 months
later from active CNS lupus, having not proceeded to transplant.The seven
remaining were free of signs of active lupus at a median follow-up of 25 months
posttransplant. The high-dose chemotherapy consisted of cyclophosphamide 200mg/kg,
methylprednisolone 1 gm and equine ATG 90mg/kg. Eligible patients had WHO
class III-IV glomerulonephritis, or lupus cerebritis or transverse myelitis,
or lupus vasculitis involving the heart or lung parenchyma - all unresponsive
to at least six cycles of intravenous cyclophosphamide; or lupus-associated,
life-threatening, severe cytopenias unresponsive to standard-dose cyclophosphamide;
or catastrophic anti-phospholipid syndrome. Of the 7 patients, 2 were referred
because of progressive lung disease and hypoxia at rest, but were not ventilator-dependent.
One patient with alveolar hemorrhage was oxygen-dependent and required intermittent
ventilator support. Four had WHO class III-IV glomerulonephritis and nephrotic
syndrome. Five had a creatine clearance rate below the normal range. All had
uncontrolled hypertension requiring four antihypertensive agents. Two had
myocardial hypokinesis. Two had seizures immediately pretransplant.
Refractory
Autoimmune Cytopenias
Refractory cytopenias represent a highly heterogeneous group. The mechanisms
appear to be more clearly defined with putative antibodies directed against
epitopes on one or more hematopoietic cell lines. Complete responses were
published very early a few patients with ITP and followed soon by failures.
A total of nine patients are included in the data base, two of them with complete
remissions . The largest series of 14 patients with ITP is from the NIH (
ref) . All had severe refractory ITP non- responsive to at least three treatment
regimens and all spenectomised. Eight showed a clear response and six a sustained
complete remission beyond two years. None died of transplant related complications.
Conditioning was with Cy and all 16 received a CD34 selected graft. The other
patients with cytopenias were heterogeneous: pure red cell aplasia ( n=4),
autoimmune haemolytic anaemia (n= 2) and Evans Syndrome ( n=1). Median age
was 31(4-45) years and median time to transplantation was 93 months ( 12-236).
Most received stem cells from blood mobilised with either growth factors (
n=7) or Cy plus growth factors ( n=6). There were two bone marrow harvests.
Conditioning regimens included Cy alone ( n=3), Cy with other drugs or ATG
( n=9), melphalan (n=2) or were fludarabine based.(n=2).Most (N=12) had purging
of immune cells from the graft product.
Conclusion
The analysis of the results of ASCT in RD lends support to a new capacity
and tool we may have in interfering with the disease evolution. New trials
are exploring now the effects of the ASCT in RD. The hope of many researchers
is now focused on the results of these trials in order to definitively state
that ASCT is an effective treatment The phase I and II data suggest that for
some highly selected AD patients ASCT could offer another treatment option
with an ethically acceptable bemefit /risk ratio. However it will only be
through multicentre , international randomised controlled trials that this
suggestion could become evidence based. The hard work starts now.
Alan
Tyndall
Dipartimento di Reumatologia Università di Basilea
Svizzera
Marco
Matucci Cerinic
Dipartimento di Medicina Sezione di Reumatologia
Università di Firenze Italia Tradotto da Interpres sas
