1. health and fitness
2. therapy

Thalidomide as Salvage Therapy for Chronic Graft-Versus-Host Disease
By Pablo M. Parker, Nelson Chao, Auayporn Nademanee, Margaret R. O‘Donnell, Gerhard M. Schmidt,
David S. Snyder, Anthony S. Stein, Eileen P. Smith, Arturo Molina, Daniel E. Stepan, Ashwin Kashyap, Ina Planas,
Ricardo Spielberger, George Somlo, Kim Margolin, K. Zwingenberger, K. Wilsman, Robert S. Negrin,
Gwynn D. Long, Joyce C. Niland, Karl G. Blume, and Stephen J. Forman
Thalidomide has been reported to be an effective agent for
treatment of chronic graft-versus-host disease (CGVHD). To
determine the efficacy of this agent in patients with refractory CGVHD a total of 80 patients who failed to respond to
prednisone (PSE)orPSE
and cyclosporine(CSA)
were
treated with thalidomide. Sixteen patients (20%) had a sustained response, 9 with a complete remission and 7 with a
partial response. Twenty-nine patients (36%) had thalidomide discontinued because of side effects, which included
sedation, constipation, neuritis, skin rash, and neutropenia.
Side effects were reversible with drug discontinuation except for mild residual neuritis in one case. Rashes and neutropenia have not previously been reported as thalidomide
side effects when used for CGVHD treatment. We conclude
thalidomide is immunosuppressive and active in the treatment of CGVHD. A high incidence of reversible side effects
limited dose intensity and reduced the number of patients
who could benefit from treatment.
0 1995 by The American Society of Hematology.
C
PUVA had complete remissions. An additional 21 patients
received thalidomide as part of primary therapy for newly
diagnosed poor-risk CGVHD. Sevenof these had a complete
response(CR).Toxicityappeared
to be minimaland was
limited to sedation and constipation, or sensory neuropathy
reversible with discontinuation of medication.
To extend the experience with this medication we have
performed a phase I1 trial of thalidomide as salvage treatment
for CGVHD refractory to primary therapy with PSE alone
or in combination with CSA. This report summarizes results
of treatment in 86 patients.
HRONIC GRAFT-VERSUS-HOST disease (CGVHD)
isthe principal cause ofmorbidity andnonrelapse
mortality for patients reaching day100 after allogeneic bone
marrow transplant (BMT). CGVHD occurs in 20% to 50%
of these patients with mortality rates ranging
from 20% to
70% depending on the associated risk factor^."^ The most
prominent risk factor for poor outcome
is progressive presentation (acute GVHD, which does not clear and evolves into
CGVHD), which is consistently demonstrated in risk analyses by different groups.”‘ Concomitant thrombocytopenia or
a combination of liver involvement with skin CGVHD with
lichenoid skin histology have also been found to carry poor
Theseanalyses reflect what has so far been
achieved with treatment with standard available medication
for CGVHD, namely prednisone (PSE), cyclosporine (CSA),
and azathioprine.
Thalidomidewas introduced as a sleeping pill butwas
removed from the market because of its teratogenic effects.
The discoverythat
thalidomide has immunosuppressive
propertiesled toitsusefor
thetreatment of CGVHD. It
was first reported in limited uncontrolled case reports with
encouraging res~lts.’.~,~
Vogelsang et alx,9have performed
the most extensivestudies on the use of this drugfor
CGVHD both in a mouse model of the disease and subsequently in a phase I1 clinical trial. In this latter trial 7 of 23
patients with CGVHD refractory to primarytherapy with
prednisone alone or with additional CSA, azathioprine, or
From the Department of Hernuto1og.y and Bone Marrow Transplantation, City of Hope National Medical Center, Duurte; the Bone
Murrow Trunsplant Program, Stanford UniversiQ Medical Center,
Stunford, CA; and Grunenthal GMBH, Stolberg, Germany.
Submitted March 23, 1995; accepted July 7, 1995.
Supported by Grants No. POI-CA-40206,POI-CA-49605, and
CA 33572 from the National Cancer Institute, Department of Health
and Human Services, Bethesda, MD.
Address reprint requests to Pablo M. Parker, MD, Department
of’ Hemutology and Bone Marrow Transplantation, City of Hope
Nutional Medical Center, 1500 E Duarte Rd, Duarte, CA 91010.
The puhlimtion costs of this article were defrayed in part by page
churge puyment. This article must therejbre be hereby marked
“advertisement” in uccordance with 18 U.S.C. section 1734 solely to
indicate this ,fact.
0 I995 by The American Sociely cf Hematology.
0006-497/~5/X609-0028$3.00/0
3604
PATIENTS AND METHODS
All patients signed informed consent forms approvedby the Institutional Review Boards of the City of Hope National Medical Center
and Stanford University. All female patients consented to take birth
control pills while on thalidomide.
Patients. BetweenSeptember 1987 andJanuary 1994, 86 paa total of 318 whodevelopedCGVHDwere
tients(27%)from
enrolledat both institutionsonthisstudy.Sixpatientswere
not
evaluable. The reasons were as follows: 4 died 5 to 13 days from
starting thalidomide from interstitial pneumonia with severe refrac= l ) ; 1 patient
tory CGVHD (n = 3), and from Candida sepsis (n
responded but CSAhadalsobeenstartedforCGVHD
1 I days
earlier; and 1 patient stopped thalidomide after 10 days of therapy
because of anxiety over possible future side effects. The characteristics of the 80 evaluable patients arelisted in Table l . All patients had
undergone allogeneic BMT from HLA identical siblingsor matched
unrelated donors for hematologic malignancies, aplastic anemia, o r
thalassemia. The preparative regimens varied according to disease
and remission status and have
been reported elsewhere.“”’ These
included fractionated total body irradiation (1,320 rad) with either
cyclophosphamide 120 m g k g or etoposide 60 mgkg, o r busulfan
16 m g k g with cyclophosphamide 120 mgkg. T depletion was not
used. Acute GVHD prophylaxis varied over time according to institutionally active protocols and included CSA and PSE,CSNmethotrexate (MTX)/PSE, and Xomazyme/CSA/PSE.’ ’,IJ All patients on
theseGVHDprotocolswerescheduled
to continue CSA with or
without PSE until day 180 after BMT.
Diagnosis. The diagnosisofCGVHD was madeaccording to
standard clinicalcriteria.’.’’.’” The diagnosis was confirmed histologicallyexcept in cases of mouth CGVHD or when liver or lung
biopsies were necessary and the responsible physician did
not feel
coagulation parameters were safe. Twenty-two patients didnot have
a biopsy to confirm the CGVHD histologically.
All patientshadextensiveCGVHD(generalizedskin,liver
histologic
involvement with hyperbilirubinemia, or advanced
Blood, Vol 86,No 9 (November l), 1995: pp 3604-3609
3605
THALIDOMIDE THERAPY FOR CGVHD
Table 1. Characteristics of Patients With Chronic GVHD
Total
Agelmedian
Diagnosis
A ML
remission
First
remission
=First
ALL
remission
First
remission
ZFirst
CML
CP
AP/BC
SAA
MDS
MPS
Thalassemia
sibling
identical HLA
MUD/MRD*
Acute GVHD prophylaxis
PSE/MTX
CSA/PSE
CSNMTWPSE
CSNPSE/Xoma
CSNPSEIATG
Grade of acute GVHD
0-11
Ill-IV
(range)
80
26/(6-50)
10
3
7
6
24
14
8
5
2
1
72
8
3
54
21
1
1
59
21
MUD, matched unrelated donor; MRD, matched related donor.
changes, or multiple organ CGVHD) according to previously published criteria.2 Thirty-eight of the 80 patients also were classified
as having poor prognostic features or high risk due to either thrombocytopenia, progressive presentation, or combined skin and liver
CGVHD.
Eligibility. Patients were eligible if they
had
progressing
CGVHD after 1 month (n = 31) or stable but unimproved CGVHD
(n = 40) after 2 months of PSE or PSElCSA treatment. Patients
who responded to PSE or PSElCSA but flaredwhen these were
tapered and required maintaining PSE at doses 220 mg/d were also
eligible (n = 7). Patients were also eligible for thalidomide at PSE
doses 5 2 0 mg/d if CGVHD was present and ongoing PSE treatment
was contraindicated due to steroid related complications (aseptic hip
necrosis, prior or concurrent fungal infections, or intolerance of
cushingoid effects) (n = 2). Prior CSA therapy was not required but
most patients had failed CSA (n = 67) as well and had thalidomide
added to their existing PSWCSA treatment or were unable to receive
further CSA because of CSA-related end-organ toxicity (n = 13)
(Table 2). The target protocol eligibility dose of PSE 20 mg/d was
chosen taking into account the long period of time these patients
need to be on therapy for CGVHD with the potential for longterm steroid related complications. The median time on therapy for
CGVHD before thalidomide was started was 4.5 months (range, 1
month to 3 '/z years).
Thalidomide. Thalidomide was obtained either from Reiza do
Nordeste (Sao Paulo, Brasil) or Grunenthal GMBH (Stolberg, Germany). Thalidomide was started at a dose of 100 mg orally four
times a day and if no side effects were encountered was escalated
to 200 and 300 mg four times a day. Drug levels were not evaluated
for thalidomide. CSA levels were also not routinely done inall
patients with CGVHD and their influence on outcome was therefore
not analyzed in this trial. Patients were maintained on PSE and CSA
when started on thalidomide. PSElCSA were then slowly tapered if
CGVHD improved with addition of thalidomide.
Responses. The frequency of follow-up for patients on thalidomide was determined by physician assesment of clinical need. Retrospective chart review was used to determine outcome. A complete
response was classified as complete disappearance of all clinical
manifestations of CGVHD. A partial response included a 50% improvement in objective parameters of CGVHD manifestations (extent of skin involvement, total bilirubin, pulmonary function tests
(PIT) in bronchiolitis obliterans with organizing pneumonia
[BOOP]). A response in oral CGVHD was more difficult to quantitate objectively but required symptomatic improvement as well as
physician assessment of comparative improvement. Patients were
considered to have no response if CGVHD progressed after 1 month
or failed to improve after 3 months of thalidomide therapy. Only
patients with sustained responses to thalidomide without reprogression were included among the responders. Patients who responded
but then had to discontinue the drug due to side effects were considered treatment failures if the disease then progressed. All patients
in this report have been on protocol more than 3 months. Survival
was analyzed from the start of thalidomide therapy.
RESULTS
Response. Eighty patients are evaluable for response and
toxicity. Their characteristics at transplant, acute GVHD proTable 2. Characteristics of Chronic GVHD at Study Entry
Total
CGVHD Site
Mouth
Skin
Skin/mouth
Liver (+/Fmouth)
Liverlskin
BOOP
GI (+skin/mouth/liver)
Thrombocytopenia
Onset after BMT
Progressive
3-6 m o
6-12 m o
12-24 m o
>24 m o
Prior treatment
None
PSE
PSE/CSA
PSE/CSNPUVA
PSE/CSA/CTX
Time from BMT to thalidomide
start
3-6 m o
6-12 rno
12-24 mo
>24 m o
Extensive CGVHD*
High risk (thrombocytopenia,
progressive CGVHD,
combined skin & liver
CGVHD)
Standard risk (all others)
80
15
9
15
15
12
9
5
29
23
20
25
10
2
1
12
62
4
1
Duration: median (range)
12mo
5 m o (1-26 mo)
4 m o (1-36 rno)
6 m o (2-36 m o )
3%yr
19
29
14
18
80
38
42
* Limited, localized skin +/- hepatic dysfunction; extensive, generalized skin, advanced liver, or multiple organ CGVHD.
3606
PARKER ET AL
Table 3. Summary of Results
~~
Response
CR
Patients
ALL CGVHD
High risk CGVHD*
Standard risk
CGVHD
Mouth only
Skin only
Skinlmouth
Liver +/- mouth
Liver/skin
Total
CR
80
9
38
4
42
15
9
15
15
12
5
PR
3
16 (20%)
6 (16%)
5
1
10 (24%)
4 (24%)
0 (0%)
6 (40%)
5 (33%)
0 10%)
1(11%)
-
1
5
5t
-
-
-
BOOP
9
GI (upper l)(lower 4)
+/- liver/mouth/
lung skin
5
-
+ PR/
Percent
7
2
-
-
MR
-
1
0 (0%)
Survival
Responders
Nonresponders
16
64
9/9 (100%) 4/7 (57%)
13/16 (81%)
30/64 (47%)
Abbreviation: MR. minor response (<PR).
* W i t h thrombocytopenia, progressive CGVHD or combined skin/
liver CGVHD.
t One mouth PR, skin no change.
tients with gastrointestinal (GI) CGVHD were treated with
thalidomide (Table 2). One had esophageal web syndrome.
The other four had lower GI CGVHD. All had additional
other organ involvement (skin/mouth/liver).
They
were
placed on thalidomide with the intent of testing if the constipating effect of the drug would allow for absorption and a
of diarrhea from
therapeuticeffect
even inthesetting
CGVHD. None of these patients became responders.
Fifteen of the 16 patients with refractory CGVHD who
responded to thalidomide had failed PSE and CSA, one responding patient with mouth only CGVHD wason PSE only
because of CSA intolerance. Among the 16 patientswho
responded to thalidomide the median time to achieve a PR
was 3 months (range 1 to 12 months). The median duration
of thalidomide therapy was 16 months (range 4 months to
4'12 years). The median thalidomide dose was 200 mg four
times daily (range, 100 mg three times a day to 400 mg four
times a day). One patient with a PR for skidmouth CGVHD
who had no side effects with the 300 mg three times a day
dose was dose escalated to 400 mg four times a day. All
patients initially remained on PSE or PSE/CSA in addition
to the thalidomide until a response was achieved and were
then sequentiallytapered off allthree drugsover several
months. The choice of drug to taper first was by physician
preference althoughPSE wasusuallytapered
first. Six of
the 16 responders remain off all immunosuppressive medications 8 to 36 months after discontinuing thalidomide. Seven
additionalrespondingpatientsremainontaperedthalidomide or PSE and/or CSA for a median of 3 years (range,
1 to S years) from thestart of thalidomide therapy. Three
additional patients with a sustained response to thalidomide
died while still on tapered doses of thalidomide, PSE/CSA
from infectious complications of CGVHD (two dueto pneumonia and one fromencephalitis) 6, 16, and 2.5 months from
the start of thalidomide treatment.
An additional eight patients were
improving on thalidomide but had to discontinue the drug becauseof side effects
and then subsequentlyprogressed.Theywere
countedas
treatment failures. Five had mouth CGVHD only, one had
combinedskidmouth,onecombinedmouth,
esophageal
web, and bronchiolitis, and one liver involvement. All flared
again after discontinuation of thalidomide. Six remain with
persistentunimproved CGVHD on PSE/CSA 1 to 5 years
after discontinuing thalidomide. Two have hada PR, one
with additional treatment, which included photopheresiswith
phylaxis regimen, and gradeof subsequent acute GVHDare
summarized in Table l . The status of their CGVHD at entry
to the thalidomide protocol is summarized in Table 2.
Sixteen of 80 patients (20%) had asustained response
(Table 3). Nine patients had a CR and seven a partial response (PR). Overall 6 (16%) of 38 patients with high-risk
CGVHD and 10(24%) of 42 with standard-risk CGVHD
respondedtothalidomide(Table
4). The high-risk group
included 2 of 23 patients with progressive CGVHD ( s k i d
mouthwith thrombocytopenia [n = l] and liver [n = l])
and 4 with thrombocytopenia with mouth (n = 1) or liver
(n = 3) involvement. None of 12 patientswith combined
skin and liver CGVHD responded.
Four of 15 patients with mouth only CGVHD and 5 of
14 with combined skin and mouth CGVHD responded. One
of these latter patients had a PR for severe sclerodermatous
skin involvement. No other patienttreated for severe sclerodermatous skin involvement (isolated or with other organ
involvement) responded. One other patient with combined
mouth and skin CGVHD had a PR for the mouth CGVHD
on thalidomide without improvement of skin manifestations.
Table 4. Toxicitv of Thalidomide
The latter then improved on additional PUVA therapy. Five
with or without associated mouth
of 15 with liver CGVHDTotal
Sedation
CGVHD (totalbilirubin median 11.6, range 4.9 to 25.1)
Stopped treatment
responded. One other patientwith a PR for mouth CGVHD
Total
Constipation/nausea
also had BOOP with PFT at 60% of pre-BMT values and
Stopped treatment
was oxygen dependentTotal
atstart of thalidomide. Her PFT only
Neuritis
Stopped treatment
improved by 20% on thalidomide and did not meet criteria
for a PR, but she became oxygen independent and Total
her exer-rash New skin
Stopped treatment
cise tolerance improved. Unfortunately, she died ofan acute
Total
Neutropenia
encephalitis of presumed
viral origin
after remaining 2 years
Stopped treatment
on thalidomide treatment. None of the other eight patients
Stoppedtreatment-alltoxicitiesTotal
with BOOP treated with thalidomide responded. Five pa-
32 (40%)
7
24 (30%)
3
3 (5%)
3
13 (16%)
6
14 (18%)
10
29 (36%)
THALIDOMIDE THERAPY FOR CGVHD
PSWCSA and the other by continuing longer treatment with
PSWCSA.
Of note one of the patients entered on this protocol had
not received any CGVHD treatment for l year before starting
thalidomide (Table 2). She had sclerodermatous skin
CGVHD with contractures and had developed a psychotic
episode after 6 months of combined PSWCSA treatment.
Both were tapered off and she remained with persistent
CGVHD off therapy until thalidomide became available 1
year later. She died 2 months after starting thalidomide from
CGVHD-related complications and was considered a treatment failure.
Toxicity. There was significant difficulty in maintaining
patients on this medication because of a high frequency of
both patient intolerance of sedation and bowel distension
and additional side effects requiring drug discontinuation.
Previous reports of side effects when using thalidomide for
CGVHD have listed sedation and constipation as the most
common with tachyphylaxis as treatment continues leading
to adequate patient tolerance? Sensory neuropathy has also
been reported as a less frequent complication but its appearance warrants taking the patient permanently off medication."
We also encountered frequent initial sedation and constipation (Table 4). However, not all patients adapted with time
and seven patients discontinued the drug due to persistent
sedation even with dose reduction. Although constipation
was less problematic resolving with stool softeners, the
bowel distension caused by thalidomide led to stopping medication in three patients because of abdominal discomfort
with nausea and vomiting. Symptoms of sensory neuropathy
led to discontinuation of medication in three patients. One
patient still has neuropathic symptoms 4 months after stopping thalidomide. Neuropathy resolved in the other two
within 2 months of stopping medication.
We also encountered two additional side effects. The first
was development of a new skin rash (in 4 patients) clinically
suggestive of CGVHD (when none had beenpreviously present) or flare of a pre-existing skin GVH rash (in 9 patients)
within 4 weeks of starting treatment. PSWCSA doses had
not yet been tapered in four (2 with and 2 without preexisting skin CGVHD) of these patients when the rash developed or flared. Skin biopsies were done in 5 of these 13
patients and showed histologic changes suggestive of
CGVHD in all cases. In all cases improvement or resolution
of skin rash occurred as PSE doses were increased at the
same time that thalidomide was held. Two patients with skin
CGVHD were rechallenged with thalidomide when the skin
rash had improved and flaredwith the second challenge.
Four other patients were not restarted on thalidomide after
developing a rash due to physician andor patient reluctance
to restart therapy due to this side effect. One patient who
was started on thalidomide for skin (localized to face and
upper torso) and mouth CGVHD developed a severe Stevens-Johnson like syndrome with erythroderma with bullous
lesions and conjunctival and mouth mucosal involvement 2
weeks after starting therapy. His PSWCSA had not yet been
tapered. He required hospitalization and high-dose steroids
with resolution of the acute skin changes but persistence of
3607
his prior CGVHD baseline skin lesions. He subsequently
died from CGVHD progression with BOOP but without new
skin lesions. Seven other patients were restarted on thalidomide after the skin rashes improved and did not reflare.None
of these became responders and were therefore eventually
takenoff
medication because of CGVHD progression.
Among the 13 patients who developed skin rashes with the
start of thalidomide treatment there was a higher incidence
of two poor risk factors. Six (46%) had prior thrombocytopenia and seven (54%) had progressive presentation of
CGVHD. This compared with the 67 patients who did not
develop skin rashes in which 23 (34%) had thrombocytopenia and 16 (24%) had progressive CGVHD. The third poor
prognostic risk factor, combined skidiver CGVHD, was not
more frequent in this group with only 1 of 13 (8%) in those
developing a skin rash as opposed to 1 1 of 67 (16%) in those
who did not.
The second additional unexpected toxicity was neutropenia (Table 4). This occurred in 14 patients at a median of
4 weeks (range 1 to 11 weeks) from starting therapy without
affecting red blood cell (RBC) and platelet counts. The median pre-thalidomide white blood cell (WBC) count was 4.2
X IO9& (range, 1.6 to 6.4) with a median WBC nadir of 0.9
X IO9& (range, 0.3 to 2.0). The median thalidomide dose
at which neutropenia developed was 200 mg four times a
day (range, 50 mg twice a day to 300 mg four times a
day). The neutropenia usually resolved within 1 month of
discontinuing thalidomide (except in three patients who died
between 3 weeks and 1 'l2months from starting thalidomide
due to refractory rapidly progressing CGVHD). As there are
other causes of neutropenia in BMT patients, especially with
CGVHD, six patients were rechallenged with thalidomide
and again became neutropenic. The neutropenia was the direct reason for discontinuing thalidomide in 10 of 14 patients. In the other four patients, one had associated neuropathy and two others discontinued use due to refractory
CGVHD. Only one patient who developed neutropenia remained on treatment. He had responded to thalidomide and
was rechallenged and developed recurrent neutropenia, but
was able to remain on 50% doses with intermittent G-CSF.
Among the 14patients who developed neutropenia 12 (86%)
had thrombocytopenia, 8 (57%) had progressive CGVHD,
and 3 (21%) had combined skidiver CGVHD. This compared with 14 (21%) with thrombocytopenia, 13 (20%) with
progressive CGVHD, and 9 (14%) with combined skidiver
CGVHD in the 66 patients who did not have neutropenia.
Survival and causes of death. Forty-three of the 80
(53%) treated patients are alive. Thirteen of sixteen (81%)
patients who responded are alive with a median follow-up
of 2 years/8 months (range, 6 months to 5 years). Thirty of 64
(47%) who did not respond survive with a median survival of
l year12 months (range, 1 month to 7 years14 months).
Thirty-seven patients died. Twenty-seven died from organ
failure with refractory CGVHD: 11 with hepatic failure, 6
with interstitial pneumonitis or respiratory failure associated
with severe sclerodermatous or liver CGVHD, 10 with bronchiolitis obliterans. An additional 5 patients died from infection: two pneumonias, one disseminated aspergillus, one
brain abscess, organism unidentified, and one encephalitis
3608
of presumed viral origin. One patient died from a cerebrovascular accident, one committed suicide3 months off thalidomide with persistent liver CGVHD. Three patients died at
home. Two were on thalidomide with severe sclerodermatous CGVHD. One was off thalidomide for 10 months with
persisting CGVHD and severe immunodeficiency with disseminated
mycobacterium
avium-intracelulare
infection.
There were no relapses in these 80 patients with refractory
CGVHD.
PARKER ET AL
we were impressed with the timing of the rash occumng
soon after starting the drug, even in patients in whom PSEl
CSAhad not yet beentapered orwho had had noskin
CGVHD before starting thalidomide. In addition, two patients who wererechallengedhadskin
flare ups asecond
time. Two brief references to skin rashes as drug side effects
in patients treated with thalidomide are noted in a review on
the use of the drug for leprosy reaction.Ix." It is noteworthy
that the patients in our study who developed skin rashes on
thalidomide had more advanced CGVHD. It is possible that
DISCUSSION
in some patients thalidomide may be initially causing adrugrelated skin reaction, which maythen be exacerbating underThese resultsconfirm thalidomidehas activity against
lying subclinical or pre-existing skin CGVHD, just ab sun
CGVHD, which has failed
treatment with PSE or PSElCSA.
exposure
has beenreported to flareskinCGVHD.'
This
Sixteen (20%) of 80 patients in this series had a sustained
would
explain
the
finding
of
histologic
GVHD
when
skin
response. When looking at prognostic subgroups more rebiopsies were performed in these cases.
sponses wereseen in patients with standard-risk CGVHD
Neutropeniahas not beenreported as a sideeffect of
(10 of 42) than in those with high-risk disease (6 of 38).
thalidomide
inthetransplantliterature.
There is a report
There was no particular
riskprofile that predictedwhich
from
the
World
Health
Organization
on
a
tendencyto neutropatients would respond except that responses were rare in
those with CGVHD, which has had poor outcome with stan- penia in leprosy patients treated with thalidomide.*" We observed a frequent association (14 of 80 patients) of neutrodard available treatment such as BOOP, combined skin and
penia with the start of thalidomide therapy. In these cases it
liverCGVHD,andprogressive
presentation. Wedidsee
was not possible to differentiate the neutropenia from that
responses in patients with thrombocytopenia as the only asoccurring as a complication of CGVHD where multiple facsociated poor prognostic factor. Mostof the responses were
tors including infection, folatehtritional deficiency. or drug
in patients with isolated mouth or liver CGVHD and skin
effects in addition to the probableautoimmune effects of
involvementthatdid
not includesevere sclerodermatous
GVHD on marrow function may have been active."." Howmanifestations. Thalidomide therapy was a new beneficial
ever, again the timing of this complication, occurring in the
treatment modality for these patients since over half had a
first weeks after thalidomide was started in patients without
CR for CGVHD, which up to that point was refractory to
prior neutropenia, and the recurrence of neutropenia on retreatment and all were able to either discontinue or significhallenge with thalidomide in six patients suggests this was
cantly reducePSE treatment requirements. An additional
a drug-related effect. Patients treated
with thalidomide for
eight (10%) patientswere improving but had to discontinue
CGVHD should have theirblood counts frequentlymonithe drug due to side effects followed
by progression. Theretored during the first weeks of therapy.
fore, they did notbenefit from treatment. Although they were
We did not find thalidomide to be useful for treatment of
notincluded in our listofresponders,their
improvement
the majority of patients with CGVHD, which was prohas significance in emphasizing that the immunosuppressive gressing after failing PSEKSA. We may have been limited
properties of thalidomide may be better than what is sugby the high incidence of side effects (especially neutropenia
gested by our limited 20% response rate, and that the diffiand skin rashes) in this group, which did not allow for dose
culty in dose escalating or maintaining patients on the mediintensity. We did not measure drug levels in our patients but
cation due to side effects may bea major problem with this
the majority did nottolerateescalatingthalidomide
doses
drug in the treatment of CGVHD.
above 400 to 600 mgld, which was below the target doses
Atotalof 29 (36%) of the 80 patients were takenoff
of 800 to 1,200 mgld. Vogelsang et al used doses of 800 to
treatment because of side effects. Ten of thesewere unableto
1,600 mgld to reachtargettherapeuticserum
levels." It is
continue treatment due tosedation or constipatiotdabdorninal
possible that if thalidomide were used earlier as part of first
distension and three due to
neuropathy. An additional16
line therapy, patients would tolerate dose intensive therapy
patients had skin rashes or neutropenia, which required disbetter with animproved response rate. Vogelsang et al'
continuing medication. Sedation, constipation, and neuropatreated patients with poor risk features with thalidomide as
thy are well-described complications of thalidomide. In a
part of first line therapy in combination with PSElCSA in
report by Vogelsang et al,9 sedation and constipation didnot
theirclinical trial. This led to anexcellentresponserate
apparentlylead to difficultyincontinuingtreatment
since
(7 of 21 patients had a CR) and improved survival (48%)
compared with historical
controls
of
However,
the
they seemed to predominate only in the first weeks of therhistorical controls received PSEorPSEhzathioprine
and
apy. Only 6 of 43 patients were withdrawn from that study
did not uniformlyreceive CSA. Sullivanet a12.'4.25found
because of drug side effects (4 because of neuropathy and
combined PSElCSA also improvedsurvival to an equivalent
2 because of noncompliance with medication).
level of 50% compared with 26% overPSE or CSAas
New skin rashes were not reported by Vogelsang et al.
single agents for CGVHD patients with poor risk defined by
New skin rashes or exacerbation of previously existing skin
thrombocytopenia. Therefore, the trueresponse and survival
CGVHD occurred in a significant number of our patients
benefit of adding thalidomide to PSElCSA as first line ther(13 of 80 or 16%) (Table4). The simplest explanationwould
apy for CGVHD needs to be studied in a randomized trial.
be that these rashes represented flare
a of CGVHD. However,
THALIDOMIDE THERAPY FOR CGVHD
An additional decision stemming
from the possibility that
earlierintroduction of thalidomide in CGVHD treatment
may improve efficacy would be to include it in
CGVHD
prophylaxis. We are currently conducting a placebo-controlled randomized trial of thalidomide added to CSA for
CGVHD prophylaxis in an adult BMT population.
REFERENCES
I. Vogelsang GB, Wagner JE: Graft-versus-host disease, in
Forman SJ ( 4 ) : Hematology/Oncology Clinics of North America,
v01 4:3. Philadelphia, PA, Saunders, 1990, p 625
2. Sullivan KM: Graft-versus-host disease, in Forman SJ, Blume
KG, Thomas ED (eds): Bone Marrow Transplantation. Boston,
Blackwell Scientific, 1994, p 353
3. Sullivan KM, Mori M, Witherspoon R, Sanders J, Appelbaum
FR, Storb R: Alternating day cyclosporine and prednisone (CSP/
PRED) treatment of chronic graft-versus-host disease (GVHD): Predictors of survival. Blood 76:568a, 1990 (abstr)
4. Wingard JR, Piantadosi S, Vogelsand GB, Farmer ER, Jabs
DA, Levin LS, Beschomer WE, Cahill RA, Miller DF, Harrison D,
Sara1 R, Santos GW: Predictors of death from chronic graft versus
host disease after bone marrow transplantation. Blood 74: 1428, 1989
5. Lim SH, McWhannell A, Vora AJ, Boughton BJ: Successful
treatment with thalidomide of acute graft-versus-host disease after
bone-marrow transplantation. Lancet 1:117, 1988
6. Heney D, Norfolk DR, Wheeldon J, Bailey CC, Lewis IJ,
Barnard DL: Thalidomide treatment for chronic graft-versus-host
disease Br J Haematol 78:23, 1991
7. Wood P, Proctor SJ: The potential use of thalidomide in the
therapy of graft-versus-host disease-A review of clinical and laboratory information. Leuk Res 14:395, 1990
8. Vogelsang GB, Hess AD, Friedman U,Santos GW: Therapy
of chronic graft-v-host disease in a rat model. Blood 74:507, 1989
9. Vogelsang GB, Evan MD, Farmer ER, Hess AD, Altamonte
V, Beschorner WE, Jabs DA, Corio RL, Levin LS, Colvin OM,
Wingard JR, Santos GW: Thalidomide for the treatment of chronic
graft-versus-host disease. N Engl J Med 326:1055, 1992
IO. Shank B, Chu F, Dinsmore R, Florence CH, Kapoor N, Kirkpatrick D, Teitelbaum H, Reid A, Bonfiglio P, Simpson L, O’Reilly
RJ: Hyperfractionated total body irradiation for bone marrow transplantation. Results in seventy leukemia patients with allogeneic
transplants. Int J Radiat Oncol Biol Phys 9:1607, 1983
1 I . Blume KG, Forman SJ, O’Donnell MR, Doroshow JH,
Krance RA, Nademanee AP, Snyder DS, Schmidt GM, Fahey JL,
Metter GE, Hill LR, Findley DO, Sniecinski IJ: Total body irradiation and high-dose etoposide: A new preparatory regimen for bone
marrow transplantation in patients with advanced hematologic malignancies. Blood 69:1015, 1987
12. Tutschka PJ, Copelan EA, Klein JP: Bonemarrow transplantation for leukemia following a new busulfan and cyclophosphamide regimen. Blood 70:1382, 1987
3609
13. Forman SJ, Blume KG, Krance RA, Miner PJ, Metter GE,
Hill LR, O’Donnell MR, Nademanee AP, Snyder DS: A prospective
randomized study of acute graft-v-host disease in 107 patients with
leukemia: Methotrexate/prednisone v cyclosporine Nprednisone.
Transplant Proc 19:2605, 1987
14. Chao NJ, Schmidt GM, Niland JC, Amylon MD, Dagis AC,
Long GD, Nademanee AP, Negrin RS, O’Donnell MR, Parker PM,
Smith EP, Snyder DS, Stein AS, Wong RM, Blume KG, Forman
SJ: Cyclosporine, methotrexate, and prednisone compared with
cyclosporine and prednisone for prophylaxis of acute graft-versushost disease N Engl J Med 329:1225, 1993
15. Saurat JH: Cutaneous manifestations of graft-versus-host disease. Int J Dermatol 4:249, 1981
16. Schubert MM, Sullivan KM: Recognition, incidence, and
management of oral graft-versus-host disease. NCI Monogr 9: 135,
I990
17. Ochonisky S , Verroust J, Bastuji-Garin S, Gherardi R, Revuz
J: Thalidomide neuropathy incidence and clinicoelectrophysiologic
findings in 42 patients. Arch Dermatol 130:66, 1994
18. Lakeman P, Smith WC: Thalidomide in leprosy reaction. Lancet 343:432, 1994
19. Grosshans MD, Genevieve I: Thalidomide therapy for inflammatory dermatoses. Int J Dermatol 23:598, 1984
20. Iver CG, Languillon I, Ramanujam G , Sundaresan T WHO
coordinated short term double blind trial with thalidomide in the
treatment of acute lepromatous reactions in male lepromatous patients. Bull Wld Hltb Org 45:719, 1971
21. Hirabayashi N: Studies on graft versus host (GvH) reactions.
I. Impairment of hemopoietic stroma in mice suffering from GvH
disease. Exp Hematol 9:101, 1981
22. Atkinson K, Nome S, Chan P, Zehnwirth B, Downs K, Biggs
J: Hemopoietic progenitor cell function after HLA-identical sibling
bone marrow transplantation: Influence of chronic graft-versus-host
disease. Int J Cell Cloning 4:203, 1986
23. Peralvo J, Baeigalupo A, Pittaluga PA, Occhini D, Van Lint
MT, Frassoni F, Nardelli E, Trasino A, Pantarotto M, Marmont AM:
Poor graft function associated with graft-versus-host disease after
allogeneic marrow transplantation. Bone Marrow Transplant 2:279,
1987
24. Sullivan KM, Witherspoon RP, Storb R, Weiden P, Flournoy
N, Dahlberg S , Deeg HJ, Sander JE, Doney KC, Appelbaum FR,
McGuffin R, McDonald GV, Meyer J, Schubert MM, Gauvrequ J,
Shulman HM, Sale GE, Anasetti C, Loughran TP, Strom S, Nims
J, Thomas ED: Prednisone and azathioprine compared with prednisone and placebo for treatment of chronic graft-v-host disease: Prognostic influence of prolonged thrombocytopenia after allogeneic marrow transplantation. Blood 72:546, 1988
25. Sullivan KM, Witherspoon RP, Storb R, Deeg HJ, Dahlberg
S, Sanders JE, Appelbaum FR, Doney KC, Weiden P, Anasetti C,
Loughran TP, Hill R, Shields A, Yee G , Shulman H, Nims J, Strom
S , Thomas ED: Alternating-day cyclosporine and prednisone for
treatment of high-risk chronic graft-v-host disease. Blood 72:555,
1988