Bone marrow vs extramedullary relapse of acute leukemia after allogeneic hematopoietic cell transplantation: Risk factors and clinical course

Abstract

A total of 118 consecutive adult patients with acute leukemia (78 AML, 36 ALL, and four acute mixed lineage leukemia) underwent allogeneic hematopoietic cell transplantation (HCT) after conditioning with BuCy (n=113) or a nonmyeloablative regimen of busulfan-fludarabine (n=5). After a median follow-up of 35.8 months (range, 6.4-91.0), 34 patients experienced at least one episode of leukemia relapse. Of 34 initial episodes, 14 (41%) occurred in extramedullary sites, with (n=8) or without (n=6) concomitant bone marrow involvement. The median time to relapse in the extramedullary sites was longer than that of relapse in bone marrow only (13.5 vs 6.1 months, P=0.046). Acute leukemia subtype and disease status at HCT showed an independent predictive value for overall relapse, as well as for extramedullary relapse with or without bone marrow involvement (Philadelphia chromosome positive acute leukemia vs low-risk AML, relative risk 22.68 (95% CI, 2.18-235.64); other than first CR vs first CR, relative risk 5.61 (95% CI, 1.80-17.51)), but not for bone marrow relapse. Our study suggests that there may be different pathogenetic mechanisms for bone marrow vs extramedullary relapse of acute leukemia after allogeneic HCT. The mode of relapse needs to be investigated in future reports of acute leukemia treated with allogeneic HCT.

Full text

Relapse of Leukaemia
Bone marrow vs extramedullary relapse of acute leukemia after allogeneic
hematopoietic cell transplantation: risk factors and clinical course
K-H Lee
1
, J-H Lee
1
, S-J Choi
1
, J-H Lee
1
, S Kim
1
, M Seol
1
, Y-S Lee
1
, W-K Kim
1
, E-J Seo
2
, C-J Park
2
,
H-S Chi
2
and J-S Lee
1
1
Division of Oncology-Hematology, Department of Medicine, College of Medicine, University of Ulsan, Asan Medical Center, Seoul,
Korea; and
2
Department of Laboratory Medicine, College of Medicine, University of Ulsan, Asan Medical Center, Seoul, Korea
Summary:
A total of 118 consecutive adult patients with acute
leukemia (78 AML, 36 ALL, and four acute mixed
lineage leukemia) underwent allogeneic hematopoietic cell
transplantation (HCT) after conditioning with BuCy
(n¼113) or a nonmyeloablative regimen of busulfan-
fludarabine (n¼5). After a median follow-up of 35.8
months (range, 6.4–91.0), 34 patients experienced at least
one episode of leukemia relapse. Of 34 initial episodes, 14
(41%) occurred in extramedullary sites, with (n¼8) or
without (n¼6) concomitant bone marrow involvement.
The median time to relapse in the extramedullary sites
was longer than that of relapse in bone marrow only (13.5
vs 6.1 months, P¼0.046). Acute leukemia subtype and
disease status at HCT showed an independent predictive
value for overall relapse, as well as for extramedullary
relapse with or without bone marrow involvement
(Philadelphia chromosome positive acute leukemia vs
low-risk AML, relative risk 22.68 (95% CI, 2.18–
235.64); other than first CR vs first CR, relative
risk 5.61 (95% CI, 1.80-17.51)), but not for bone marrow
relapse. Our study suggests that there may be
different pathogenetic mechanisms for bone marrow vs
extramedullary relapse of acute leukemia after allo-
geneic HCT. The mode of relapse needs to be investi-
gated in future reports of acute leukemia treated with
allogeneic HCT.
Bone Marrow Transplantation (2003) 32, 835–842.
doi:10.1038/sj.bmt.1704223
Keywords: extramedullary relapse; acute leukemia; allo-
geneic HCT
Allogeneic hematopoietic cell transplantation (HCT) is
now considered part of a standard treatment modality for a
significant subset of patients with acute leukemia.
1–3
The
curative effect of allogeneic HCT is attributable to its
ability to decrease leukemia relapse significantly when
compared to conventional or high-dose chemotherapy
including autologous HCT. This remarkable effect is due
to the graft-versus-leukemia (GVL) effect that occurs after
allogeneic HCT. Although the overall frequency of acute
leukemia relapse is less after allogeneic HCT, a high
proportion of extramedullary relapses
4–7
in extremely
diverse sites has been reported, including the brain,
8–11
head and neck,
8–10,12
gastrointestinal tract,
13
breast,
14,15
liver,
9
pancreas,
16
urogenital tract,
13,17
spinal canal and
paravertebral tissue,
8
bone and periosseous tissue,
13,14,18
pleura,
14
pericardium,
9
peritoneum,
19
and skin.
20
The
median time from allogeneic HCT to acute leukemia
relapse was longer in cases with extramedullary relapse
with or without bone marrow involvement when compared
to cases of bone marrow only relapse.
4–6
Uneven effective-
ness of the GVL effect in the body of patients was
suggested as one of the several possible mechanisms for the
increased frequency and wide distribution of extramedul-
lary relapse after allogeneic HCT.
4,6
However, other
factors, such as nature of the leukemic blasts, status of
acute leukemia at HCT, and conditioning regimen may also
influence the frequency of extramedullary relapse.
This study was undertaken to investigate (1) the
frequency of extramedullary relapse in a series of 118
consecutive patients with acute leukemia who underwent
allogeneic HCT in a single institution; and (2) whether
there are differences between relapse in the bone marrow
and extramedullary sites in terms of risk factors and
subsequent clinical courses.
Patients and methods
Patients and transplantation procedure
Between March 1995 and April 2002, 118 consecutive
patients with acute leukemia underwent allogeneic
HCT at the Asan Medical Center, University of Ulsan,
in Seoul, Korea, and were included in the study. The
initial 50 patients who were transplanted up to
December 1998 have been described previously.
6
Written
informed consents for hematopoietic cell collection and
transplantation were obtained from patients and
donors. Before July 1999, all patients received a BuCy
Received 7 February 2003; accepted 16 April 2003
Correspondence: Dr K-H Lee, Department of Medicine, Asan Medical
Center, 388-1 Poongnap-dong, Songpa-ku, Seoul 138-040, Korea.
E-mail: khlee2@amc.seoul.kr
Bone Marrow Transplantation (2003) 32, 835–842
&2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00
www.nature.com/bmt

regimen (busulfan 4 mg/kg/day orally on days 7to4
and cyclophosphamide 60 mg/kg/day by intravenous
infusion on days 3to2)
6,21
for conditioning therapy
with subsequent infusion of donor marrow cells on
day 0. Beginning in July 1999, a nonmyeloablative
conditioning regimen of Bu-Flu-ATG (busulfan 4 mg/kg/
day orally on days 7 and 6, fludarabine (Fludara
s
,
Berlex Laboratories, Richmond, CA, USA) 30 mg/m
2
/day
by intravenous infusion on days 7to2, antithymocyte
globulin (Atgam
s
) 20 mg/kg/day by intravenous infusion
on days 5 to 2, and methylprednisolone 2 mg/kg/day by
intravenous infusion on days 5 to 2)
22
was given to
patients who were elderly or who had comorbidity.
HLA matching for donor selection was based on serologic
typings for HLA-A, B, and C antigens and a molecular
typing for HLA-DRB1 antigen. For unrelated donors,
six antigens of HLA-A, B, and DRB1 were
considered. Patients who were conditioned with the
Bu-Flu-ATG regimen received granulocyte colony-stimu-
lating factor-mobilized peripheral blood hematopoietic
cells (G-CSF, 10 mg/kg/day subcutaneously for 4 days)
on days 0 and 1. None of the hematopoietic cell grafts
were T-cell depleted. The regimen for the prophylaxis
of GVHD, consisted of cyclosporine 1.5 mg/kg by
intravenous infusion every 12 h starting on day 1, which
was then switched to an oral dose when oral intake became
feasible. In addition to cyclosporine, patients received a
short course of methotrexate. Patients who were included in
a randomized trial comparing cyclosporine plus metho-
trexate vs cyclosporine alone for GVHD prophylaxis and
who were assigned to the cyclosporine alone arm and the
initial three patients who received Bu-Flu-ATG for
conditioning did not receive methotrexate. All patients
received a daily dose of G-CSF 450 mg intravenously
starting on day 0 or day 5 of infusion of donor
hematopoietic cells until peripheral blood absolute neu-
trophil count (ANC) was over 3000 ml. The same dose of
G-CSF was administered daily when ANC decreased below
1000 ml.
All patients were monitored prospectively for the
occurrence of adverse events, including GVHD,
regimen-related toxicities, and infections. Acute and
chronic GVHD were classified according to the criteria of
Przepiorka et al
23
and Sullivan et al,
24
respectively. A
diagnosis of veno-occlusive disease of the liver (VOD) was
made according to the clinical criteria of McDonald et al.
25
The severity of VOD was classified into mild, moderate,
or severe.
26
Beginning in December, 1997, hematopoietic chimerism
was analyzed from peripheral blood samples monthly
for 3 months, then every 3 months for 2 years after
allogeneic HCT. Polymerase chain reaction (PCR)-based
procedure utilizing short tandem repeats of DNA was used.
Detailed methods of PCR analysis have been described
previously.
27
Statistics and data analysis
The time to leukemia relapse was defined as the interval
between HCT and the relapse of leukemia. Median times
to relapse between the bone marrow only relapse and the
extramedullary relapse with or without bone marrow
involvement were compared using the Mann–Whitney
Utest.
Cumulative incidence curves for leukemia-relapse were
plotted using the Kaplan–Meier method and were com-
pared using a log-rank test. For calculation of overall
relapse of leukemia, relapse at any site (bone marrow,
extramedullary, or both) was considered as an event. For
calculation of the time to relapse in the bone marrow only,
patients who relapsed in the extramedullary sites with or
without bone marrow involvement were censored at the
time of relapse. On the other hand, for calculation of the
time to relapse in the extramedullary sites with or without
bone marrow involvement, patients who relapsed in the
bone marrow only were censored at the time of bone
marrow relapse. Patients who died due to complications of
HCT without a relapse of leukemia were censored at the
time of death. For the multivariate analysis, the Cox
proportional hazards regression model was used to
evaluate the predictive power of various combinations of
variables.
Post relapse survival in a patient who relapsed with
acute leukemia after allogeneic HCT was defined as the
interval between the relapse of leukemia and the last
follow-up. Post relapse survival curves were plotted using
the Kaplan–Meier method and were compared using a log-
rank test.
Results
Patients
Table 1 shows the characteristics of the 118 patients who
were included in the study. There were 78 with acute
myeloid leukemia (AML), 36 with acute lymphoblastic
leukemia (ALL), and four with acute mixed lineage
leukemia. The median ages of the patients were 33.5 years
(range, 16–57) for AML, 30 years (range, 16–57) for ALL,
and 26.5 years (range 20–38) for acute mixed lineage
leukemia. Of the 78 patients with AML, 26 had
normal cytogenetics, nine t(8;21), six t(15;17), and two
inv(16) at the time of diagnosis. When we assigned the
patients with AML according to the Southwest Oncology
Group (SWOG) criteria,
28
17 showed favorable, 28
intermediate, 15 unfavorable, and 10 undetermined risk
status. Of the 36 patients with ALL, 26 patients showed
pre-B cell type and eight T-cell type on immunopheno-
typing. A total of 10 patients showed the Philadelphia
chromosome (Ph) (eight ALL and two acute mixed lineage
leukemia). In all, 65 of the 78 AML patients, 23 of the 36
ALL patients, and two of the four acute mixed lineage
leukemia patients were in their first complete remission
(CR) status at the time of HCT. A majority of patients
received the BuCy regimen for conditioning. Five (four
AML and one ALL) patients received the nonmyeloabla-
tive regimen of Bu-Flu-ATG for conditioning. A total of 27
patients (13 AML, 12 ALL, and two acute mixed lineage
leukemia) received hematopoietic cells from unrelated
donors.
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
836
Bone Marrow Transplantation

Table 1 Patient and transplant characteristics
Characteristics AML (n ¼78) ALL (n ¼36) AmixL (n ¼4)
Age (years), median (range) 33.5 (16–57) 30 (16–57) 26.5 (20–38)
Sex
Male 39 19 2
Female 39 17 2
WBC count at diagnosis (/ml)
Less than 4000 22 8 1
4 000–30 000 29 20 1
Over 30 000 27 8 2
Peripheral blood % blast at diagnosis
Less than 25 28 11 1
25–75 27 16 2
Over 75 23 9 1
Extramedullary disease at diagnosis 0 2 0
FAB or immunophenotypic classification
M0: 5 Pre-B cell: 26
M1: 17 T cell: 8
M2: 31 Unknown: 2
M3: 6
M4: 15
M5: 2
M6: 2
Cytogenetic findings Not done: 8 Not done: 3 Normal: 1
Normal: 26 Normal: 11 t(9;22): 2
inv 16: 2 t(9;22): 8 t(9;11): 1
t(8;21): 9 del (5q): 2
t(15;17): 6 del (7q): 1
+8 or +6: 2 –7: 1
del (9q): 3 t(1;19): 1
abn 11q: 3 t(14;14): 1
del (5q): 1 Hyperdiploid: 2
t(6;9): 3 Others: 6
Others: 15
Interval from diagnosis to HCT (months)
Less than 4 40 8 2
4–12 33 19 2
over 12 5 9
Disease status at HCT
Untreated 1
Refractory 1 1
First CR 65 23 2
First relapse 4 3 1
Second CR 5 6
4Second CR 2 4
Conditioning regimen/hematopoietic cell source
BuCy/bone marrow 74 35 4
BuFluATG/peripheral blood 4 1 0
Donor age (years), median (range) 32.5 (13–55) 30 (16–55) 27.5 (26–53)
Donor sex
Male 40 24 4
Female 38 12
Type of HCT
Sibling, HLA full match 65 24 2
Unrelated, HLA match 12 10 1
Unrelated, one locus mismatch 1 2 1
Bone marrow cell dose, median (range)
Mononuclear cells ( 10
8
/kg) 0.92 (0.30–3.10) 0.85 (0.32–2.15) 0.74 (0.60–1.14)
CD34
+
cells ( 10
6
/kg) 3.40 (0.01–15.0) 2.48 (0–13.93) 2.68 (0.89–6.60)
GVHD prophylaxis
CSA alone 14 6 0
CSA+MTX 64 30 4
AmixL ¼acute mixed lineage leukemia; CSA ¼cyclosporine; MTX¼methotrexate.
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
837
Bone Marrow Transplantation

Frequency of extramedullary relapse with or without bone
marrow involvement
After the median follow-up time of 35.8 months (range,
6.4–91.0) in surviving patients, 34 of 118 patients experi-
enced at least one episode of acute leukemia relapse. Of
those 34 initial episodes of acute leukemia relapse, 14
(41%) were in extramedullary sites with or without
concomitant bone marrow involvement (Table 2). Six
patients (UPNs 14, 18, 62, 125, 159, and 204) showed
relapses in extramedullary sites only, without bone marrow
involvement. The involved extramedullary sites were the
brain (pons and peri-leptomeningeal cortex), uterus/urinary
bladder, skin, spinal canal (intradural and epidural masses,
L3-S1), breast, and soft tissue (popliteal fossa). Eight
patients (UPNs 23, 59, 67, 71, 73, 93, 118, and 165) relapsed
in extramedullary sites (bones, peri-osseous soft tissue,
intestine, mediastinum/pleura, breast, pancreas, leptome-
ninges, and skin) along with bone marrow involvement. A
total of 20 patients (59%) relapsed in the bone marrow
only. The median time to relapse in the bone marrow only
was 6.1 months (range, 2.0–83.6), while the median time to
relapse in the extramedullary sites with or without bone
marrow involvement was 13.5 months (range, 4.9–49.9,
P¼0.046, Figure 1).
Risk factors for overall relapse, relapse in the bone marrow
only, and extramedullary relapse with or without bone
marrow involvement, univariate analysis
The variables considered for univariate analysis included
age, sex, diagnosis (AML vs ALL), acute leukemia subtype,
white blood cell count at diagnosis, extramedullary disease
at diagnosis, interval from diagnosis to HCT, disease status
at HCT, conditioning regimen, donor age, sex mismatched
HCT, sibling vs unrelated donor HCT, GVHD prophy-
laxis, chimerism at 1 month, and acute and chronic GVHD.
The variables that showed significant correlation with
overall relapse of acute leukemia were acute leukemia
subtype (P¼0.028) and disease status at HCT (Po0.001).
The subtype that showed the highest rate of leukemia
relapse was Ph-positive acute leukemia (5/10, 5-year
cumulative incidence of 100%), followed by high-risk
AML (6/15, 58%). Pre-B-cell ALL without Ph and T-cell
ALL showed similar rates of relapse (7/18, 52% and 3/8,
52%, respectively). Low-risk AML showed the lowest rate
of relapse (3/17, 21%). As per disease status at HCT, 21 of
90 (35%) patients who were transplanted during their first
CR relapsed, while 13 of 28 (68%) patients who were
transplanted during other status of the disease relapsed.
Other variables that showed a trend for correlation with
overall relapse of acute leukemia after allogeneic HCT were
the diagnosis of AML vs ALL (31 vs 57%, P¼0.091),
GVHD prophylaxis with CSA only vs CSA plus metho-
trexate (16 vs 44%, P¼0.205), acute GVHD (none vs
grades I–II vs grades III–IV; 49 vs 17 vs 20%, P¼0.215),
and chronic GVHD (none vs limited vs extensive; 42 vs 44
vs 21%, P¼0.199).
Further univariate analysis was performed with separate
consideration of relapse in the bone marrow only (n¼20)
and extramedullary relapse with or without bone marrow
involvement (n¼14). Acute leukemia subtype (P¼0.005)
and disease status at HCT (Po0.001) showed significant
correlation with extramedullary relapse with or without
bone marrow involvement. The subtype that showed the
highest rate of extramedullay relapse was Ph-positive acute
leukemia (3/10, 5-year cumulative incidence of 100%),
followed by high-risk AML (4/15, 50%). Low-risk AML
showed the lowest rate of extramedullary relapse (1/17,
9%). Seven of 90 patients (17%) who were transplanted
during their first CR experienced extramedullary relapse
while seven of 28 patients (51%) who were transplanted
during other status of leukemia experienced extramedullary
relapse. On the other hand, none of the variables showed
significant correlation with relapse in the bone marrow only.
Risk factors for overall relapse, relapse in the bone marrow
only, and extramedullary relapse with or without bone
marrow involvement, multivariate analysis
The variables that showed a significance level of P-value
o0.2 on univariate analysis, as well as the patient’s age
and sex were considered in the variable selection process.
For overall acute leukemia relapse, acute leukemia subtype
(Ph-positive acute leukemia vs low-risk AML, relative risk
13.00 (95% CI, 2.43–69.63)) and disease status at HCT
(other than first CR vs first CR, relative risk 3.03 (95% CI,
1.32–6.98)) were independent variables that showed sig-
nificant correlation. The same variables showed significant
independent correlation with extramedullary relapse, that
is, acute leukemia subtype (Ph-positive acute leukemia vs
low-risk AML, relative risk 22.68 (95% CI, 2.18–235.64))
and disease status at HCT (other than first CR vs first CR,
relative risk 5.61 (95% CI, 1.80–17.51)). None of the
variables showed independent correlation with relapse in
the bone marrow only.
Clinical course after relapse of acute leukemia after
allogeneic HCT
Prognosis of the patients who relapsed after allogeneic
HCT was poor, with a median survival time of 7.8 months
(range, 0.6–58.8 þ) after the relapse of acute leukemia.
Seven patients are surviving without evidence of acute
leukemia 7.8–58.8 months after relapse.
A total of 14 patients who experienced extramedullary
relapse with or without bone marrow involvement were
treated with various modalities (Table 2) including
combination chemotherapy with subsequent donor leuko-
cyte infusion (DLI) and local radiation therapy (six
patients), combination chemotherapy with DLI (three
patients), local radiation therapy (two patients), oral
administration of imatinib mesylate (one patient), cyclo-
sporine withdrawal (one patient), and supportive care (one
patient). Two patients (UPNs 67 and 125) are surviving
without evidence of acute leukemia 15.6 and 37.8 months
after relapse.
In all, 14 of 20 patients who experienced acute leukemia
relapse in the bone marrow only were treated with
combination chemotherapy with subsequent DLI. A total
of 10 patients achieved CR. Of those 10 patients, four are
surviving without acute leukemia relapse 7.8–58.8 months
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
838
Bone Marrow Transplantation

Table 2 Summary of clinical courses of 14 patients with acute leukemia who relapsed at the extramedullary sites with or without bone marrow involvement after allogeneic HCT
UPN Age/
sex
Dx, cytogenetics Status
at
HCT
Donor Post HCT clinical courses before
relapse
Mo from
HCT
to relapse
Sites of relapse Treatment after
relapse (outcome)
Current status (post
HCT months)
VOD aGVHD GVHD
14 32/M L2, pre-B t(9;22) CR1 27/F, HLA iden
sibling
Sev None None 49.9 CNS (right lower
lobe of pons)
Whole brain
irradiation, 2000
cGy (PR)
Alive (60.6)
18 40/F M4, inv (16) CR1 47/F, HLA iden
sibling
Mod None Lim 26.9 Huge mass involving
uterus and bladder
wall
Supportive care Died with leukemia
(27.7)
23 17/F M2, normal CR1 20/M, HLA iden
sibling
None None Ext 34.4 Blood/marrow+
right mandible and
soft tissue
Chemotx+DLI+
local XRT, 4000
cGy (CR)
Died with leukemia
(62.9)
59 21/M M1, del (9q) CR1 18/M, HLA iden
sibling
Mild None None 17.0 Blood/marrow+
intestines
Chemotx+DLI
(persistent)
Died with leukemia
(17.6)
62 35/F M4, del (1p),
t(11;17)
CR1 37/F, HLA iden
sibling
None None None 5.0 Skin Cyclosporine
withdrawal
(persistent)
Died with leukemia
(14.4)
67 19/F M1, dup (3q) CR2 23/F, HLA iden
sibling
Mild Gr I None 16.9 Blood/marrow+skin
and bones (ulna,
femur)
Chemotx+DLI+
local XRT (CR)
Alive and NED
(54.7)
71 36/M L2, T cell,
trisomy 21
CR1 40/M, HLA iden
sibling
Mod None None 12.3 Blood/marrow+
mediastinum/pleura
Chemotx+DLI+
local XRT 2200
cGy (CR)
Died with leukemia
(18.4)
73 20/M L1, pre-B t(9;22) CR1 22/F, HLA iden
sibling
None None Ext 11.8 Blood/marrow+
left distal radius
with soft tissue
Chemotx+DLI+
local XRT, 3000
cGy (CR)
Died with leukemia
(16.9)
93 31/F L2, pre-B
complex ab
CR2 28/F, HLA iden
sibling
Mod None None 14.6 Blood/marrow+
left breast, pancreas,
multiple bones
Chemotx+DLI+
local XRT (CR)
Died with sepsis and
GVHD (28.3)
118 21/M L1, pre-B
complex ab
REL2 26/M, mat unrel
donor
Mod Gr III Ext 9.3 Blood/marrow+
CNS (leptomeninges)
Chemotx+DLI+
IT MTX (not
evaluable)
Died with sepsis
(11.7)
125 18/M M1, +22, CR2 23/M, mat unrel
donor
None None None 20.1 Intradural and
epidural masses,
L3-S1
Local XRT, 2000
cGy (CR)
Alive and NED
(35.8)
159 23/F M2, +6, REL1 30/F, HLA iden
sibling
None None None 8.6 Both breasts, skin,
right popliteal fossa
Chemotx+DLI+
local XRT, 3000
cGy (CR)
Died with leukemia
(20.4)
165 42/F M4, complex ab REL1 26/M, mat unrel
donor
None None None 4.9 Blood/marrow+skin Chemotx+DLI
(not evaluable)
Died with sepsis (5.9)
204 28/M ALL, pre-B
t(9;22)
REL2 35/M, HLA iden
sibling
None None None 9.4 CNS (peri-leptome-
ningeal cortex)
Gleevec for 61 days
(persistent)
Alive with CNS
leukemia (15.1)
CR1 ¼first complete remission; CR2 ¼second complete remission; REL1 ¼first relapse; REL2 ¼second relapse; HLA iden sibling ¼HLA identical sibling; mat unrel donor ¼HLA matched unrelated donor;
XRT ¼radiation therapy; sev ¼severe; gr ¼grade; mod ¼moderate; lim¼limited; ext ¼extensive; Chemotx ¼chemotherapy; DLI ¼donor leukocyte infusion; NED ¼no evidence of disease.
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
839
Bone Marrow Transplantation

after relapse. Five patients were treated with combination
chemotherapy. Of two patients who achieved CR after
combination chemotherapy, one is surviving without
evidence of leukemia 10.8 months after relapse.
The results of univariate analysis showed that patient age
(less than 25 years vs 25–40 vs over 40; 16 vs 28 vs 0%
surviving, Po0.001) and interval between HCT and acute
leukemia relapse (less than 6 months vs 6–18 vs over 18; 8 vs
17 vs 36% surviving, P¼0.023) were significant variables
that predicted post relapse survival of the patients. There
was no significant difference in post relapse survival
between patients with bone marrow only relapse and
patients with extramedullary relapse with or without bone
marrow involvement.
Discussion
The frequency of extramedullary relapse of acute leukemia
after allogeneic HCT reported in the literature varies
widely. Retrospective data from a registry reported 20
patients among 3071 patients with AML (0.7%) as having
extramedullary relapse.
8
In our series of 118 patients with
acute leukemia who underwent allogeneic HCT, 14 patients
(5-year cumulative incidence of 23%) experienced an initial
relapse involving extramedullary sites with or without bone
marrow involvement. Extramedullary relapse with or
without bone marrow involvement accounted for 41%
(14/34) of overall initial relapse. Our results are consistent
with other recent data that reported an overall frequency of
extramedullary relapse ranging from 5 to 12% of all
patients
7,20
and the percentage of extramedullary relapse
with or without concomitant relapse in the bone marrow
among the overall relapse of 27–50%.
5–7
The discrepancy of
reported frequency of extramedullary relapse may be due to
under-reporting in retrospective registry data and the
longer survival of patients in recent series due to less
patients dying of transplantation-related causes.
Factors that are known to predict relapse of acute
leukemia after allogeneic HCT include disease status at the
time of transplant, T-cell depletion of the hematopoietic
cell graft, presence of GVHD, and leukemia subtype.
1–3,28–
30
When we analyzed various pre and post transplant
variables in our series of patients, acute leukemia subtype
and disease status at HCT were significant independent
prognostic factors for overall relapse of acute leukemia.
Acute and chronic GVHD failed to show significant
correlation, although patients without GVHD tended to
show a higher relapse rate. Further analysis with separate
consideration of the different mode of relapse (bone
marrow vs extramedullary) showed that acute leukemia
subtype and disease status at HCT showed significant
correlation with extramedullary relapse with or without
bone marrow involvement, but not with bone marrow only
relapse. This finding suggests that known prognostic
factors for acute leukemia relapse after allogeneic HCT
affect bone marrow relapse and extramedullary relapse to
different degrees. Furthermore, the findings in our study, as
well as those of others,
4,5
which showed that the median
time to relapse of acute leukemia in the bone marrow was
shorter than that of extramedullary relapse with or without
bone marrow involvement suggest that different patho-
genetic mechanisms play roles in relapses in the bone
marrow and in the extramedullary sites.
The high extramedullary relapse rate of acute leukemia
in our study may be related to the low plasma level and
poor tissue penetration of orally administered busulfan
31,32
that was used in our study. One randomized study reported
a higher incidence of relapse of AML among patients who
received the BuCy regimen as opposed to the cyclopho-
sphamide-total body irradiation (TBI) regimen,
33
while
another study failed to show such a difference.
34
The nature
of relapse of leukemia was not stated in either study.
Further prospective studies with larger number of
patients and separate end points for bone marrow vs
extramedullary relapse are needed to ascertain the relative
effects of the different conditioning regimens (TBI contain-
ing vs non-TBI containing; oral busulfan vs intravenous
busulfan;
35
and myeloablative vs nonmyeloablative con-
ditioning regimen), the different degree of GVL/GVHD
effects, and the characteristics of the leukemic cells in the
different modes of relapse of acute leukemia after
allogeneic HCT. Eventually, different therapeutic strategies
for the prevention of different modes of relapse of acute
leukemia after allogeneic HCT may become apparent.
Once acute leukemia relapse occurs after allogeneic
HCT, the prognosis is poor, with a median post relapse
survival time of 7.8 months in our series of patients. The
age of patients and the interval between HCT and relapse
were the significant prognostic factors. The number of
patients in our study was too small to detect other
potentially important variables that might have prognostic
significance (i.e., mode of relapse after allogeneic HCT).
In conclusion, our study showed a high frequency of
extramedullary relapse of acute leukemia after allogeneic
HCT after conditioning with a non-TBI containing
preparative regimen (41% of all initial relapses). Extra-
medullary relapse occurred later than bone marrow relapse.
Acute leukemia subtype and disease status at HCT were
A
B
0 12243648
(83.60)
Month
Figure 1 Times to relapse of acute leukemia after allogeneic HCT were
plotted for relapse in the bone marrow (a,n¼20) and extramedullary
relapse with or without bone marrow involvement (b,n¼14). Noteworthy,
the median time to relapse in the bone marrow was shorter than the median
time to relapse in the extramedullary sites with or without bone marrow
involvement (6.1 vs 13.5 months, P¼0.046).
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
840
Bone Marrow Transplantation

significant independent predictors of extramedullary re-
lapse but not of bone marrow relapse. We propose a
separate consideration of different modes of relapse of
acute leukemia for future studies reporting the results of
acute leukemia treated with allogeneic HCT.
References
1 Giralt SA, Champlin RE. Leukemia relapse after allogeneic
bone marrow transplantation: a review. Blood 1994; 84:
3603–3612.
2 Zittoun RA, Mandelli F, Willemze R et al. Autologous or
allogeneic bone marrow transplantation compared with
intensive chemotherapy in acute myelogenous leukemia. N
Engl J Med 1995; 332: 217–223.
3 Horowitz MM, Messerer D, Hoelzer D et al. Chemotherapy
compared with bone marrow transplantation for adults with
acute lymphoblastic leukemia in first remission. Ann Intern
Med 1991; 115: 13–18.
4 Au WY, Kwong YL, Lie AKW et al. Review article. Extra-
medullary relapse of leukemia following allogeneic bone
marrow transplantation. Hematol Oncol 1999; 17: 45–52.
5 Chong G, Byrnes G, Szer J, Grigg A. Extramedullary relapse
after allogeneic bone marrow transplantation for haemato-
logical malignancy. Bone Marrow Transplant 2000; 26:
1011–1025.
6 Lee K-H, Lee J-H, Kim S et al. High frequency of
extramedullary relapse of acute leukemia after allogeneic bone
marrow transplantation. Bone Marrow Transplant 2000; 26:
147–152.
7 Simpson DR, Nevill TJ, Shepherd JD et al. High incidence of
extramedullary relapse of AML after busulfan/cyclophospha-
mide conditioning and allogeneic stem cell transplantation.
Bone Marrow Transplant 1998; 22: 259–264.
8Be
´ka
´ssy AN, Hermans J, Gorin NC, Gratwohl A. Granulo-
cytic sarcoma after allogeneic bone marrow transplantation: a
retrospective European multicenter survey. Bone Marrow
Transplant 1996; 17: 801–808.
9DelaCa
´mara R, Figuera A, Steegmann JL et al. Allogeneic
bone marrow transplantation for high risk acute lymphoblastic
leukemia. Results from a single institution. Bone Marrow
Transplant 1992; 9: 433–438.
10 Oliff A, Ramu N-P, Poplack D. Leukemic relapse 5 1/2 years
after allogeneic bone marrow transplantation. Blood 1978; 52:
281–284.
11 Lawson SE, Darbyshire PJ. Use of donor lymphocytes in
extramedullary relapse of childhood acute lymphoblastic
leukaemia following bone marrow transplantation. Bone
Marrow Transplant 1998; 22: 829–830.
12 Colombini A, Barzaghi A, Castagneto M et al. Retro-orbital
late relapse in a child with leukemia after allogeneic bone
marrow transplantation. Acta Haematol 1995; 94: 44–47.
13 Fleming DR, Greenwood MF, Garrison J et al. Lymphocyte
infusion for delayed extramedullary relapse of acute leukemia
following bone marrow transplantation. Leuk Lymphoma
1996; 21: 525–528.
14 Scha
¨fer H, Bader P, Kaiserling E et al. Extramedullary
relapses at uncommon sites after allogeneic stem cell trans-
plantation. Bone Marrow Transplant 2000; 26: 1133–1134.
15 Kelch BP, Bulova SI, Crilley P et al. An unusual
extramedullary relapse of acute nonlymphocytic leukemia
after allogeneic bone marrow transplantation. Am J Clin
Oncol 1990; 13: 238–243.
16 Au WY, Lie AKW, Liang R, Kwong YL. Isolated extra-
medullary relapse of acute lymphoblastic leukemia after
allogeneic bone marrow transplantation. Bone Marrow Trans-
plant 1999; 24: 1137–1140.
17 Forrest DL, Dalal BI, Naiman SC et al. Testicular relapse of
acute promyelocytic leukemia after allogeneic BMT. Bone
Marrow Transplant 1997; 20: 689–690.
18 Ueda S, Kanamori H, Sasaki S et al. Isolated extramedullary
relapse in knee joint after allogeneic bone marrow transplanta-
tion for Ph ALL. Bone Marrow Transplant 1998; 21: 319–321.
19 Riseberg DA, Mulvey KP, Dahut WL et al. Late extra-
medullary relapse following bone marrow transplant for AML
presenting as acute renal failure and leukemic ascites. Bone
Marrow Transplant 1994; 14: 1009–1010.
20 Michel G, Boulad F, Small TN et al. Risk of extramedullary
relapse following allogeneic bone marrow transplantation for
acute myelogenous leukemia with leukemia cutis. Bone
Marrow Transplant 1997; 20: 107–112.
21 Lee J-H, Lee K-H, Lee J-H et al. Plasminogen activator
inhibitor-1 is an independent diagnostic marker as well as
severity predictor of hepatic veno-occlusive disease after
allogeneic bone marrow transplantation in adults conditioned
with busulfan and cyclophosphamide. Br J Hematol 2002; 118:
1087–1094.
22 Lee K-H, Lee J-H, Lee J-H et al. Non-myeloablative
conditioning regimen of fludarabine, busulfan, anti-thymocyte
globulin, and methylprednisolone for allogeneic peripheral
blood hematopoietic cell transplantation. Haematologica 2001;
86: 999–1001.
23 Przepiorka D, Weisdorf D, Martin P et al. Meeting report.
Consensus conference on acute GVHD grading. Bone Marrow
Transplant 1995; 15: 825–828.
24 Sullivan KM. Acute and chronic graft-versus-host disease in
man. Int J Cell Cloning 1986; 4: 42–93.
25 McDonald GB, Hinds MS, Fisher LD et al. Veno-occlusive
disease of the liver and multiorgan failure after bone marrow
transplantation: a cohort study of 355 patients. Ann Intern
Med 1993; 118: 255–267.
26 Bearman SI, Anderson GL, Mori M et al. Venoocclusive
disease of the liver: development of a model for predicting fatal
outcome after marrow transplantation. J Clin Oncol 1993; 11:
1729–1736.
27 Choi S-J, Lee K-H, Lee J-H et al. Prognostic value of
hematopoietic chimerism in patients with acute leukemia after
allogeneic bone marrow transplantation: a prospective study.
Bone Marrow Transplant 2000; 26: 327–332.
28 Slovak ML, Kopecky KJ, Cassileth PA et al. Karyotypic
analysis predicts outcome of preremission and postremission
therapy in adults acute myeloid leukemia: a Southwest
Oncology Group/Eastern Cooperative Oncology Group study.
Blood 2000; 96: 4075–4083.
29 Lee S, Kim D-W, Kim Y-J et al. Influence of karyotype
on outcome of allogeneic bone marrow transplantation
for adults with precursor B-lineage acute lymphoblastic
leukaemia in first or second remission. Br J Hematol 2002;
117: 109–118.
30 Cornelissen JJ, Carston M, Kollman C et al. Unrelated
marrow transplantation for adult patients with poor-risk acute
lymphoblastic leukemia: strong graft-versus-leukemia effect
and risk factors determining outcome. Blood 2001; 97:
1572–1577.
31 Hassan M, Ljungman P, Bolme P et al. Busulfan bioavail-
ability. Blood 1994; 84: 2144–2150.
32 Vassal G, Koscielny S, Challine D et al. Busulfan disposition
and hepatic veno-occlusive disease in children undergoing
bone marrow transplantation. Cancer Chemother Pharmacol
1996; 37: 247–253.
33 Blaise D, Maraninchi D, Archimbaud E et al. Allogeneic bone
marrow transplantation for acute acute myeloid leukemia in
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
841
Bone Marrow Transplantation

first remission: a randomized trial of a busulfan–cytoxan versus
cytoxan–total body irradiation as preparative regimen: a
report from the Groupe d’Etudes de la Greffe de Moelle
Osseuse. Blood 1992; 79: 2578–2582.
34 Ringde
´n O, Ruutu T, Remberger M et al. A randomized trial
comparing busulfan with total body irradiation as condition-
ing in allogeneic marrow transplant recipients with leukemia: a
report for the Nordic Bone Marrow Transplantation Group.
Blood 1994; 83: 2723–2730.
35 Andersson B, Madden T, Tran HT et al. Acute safety and
pharmacokinetics of intravenous busulfan when used with oral
busulfan and cyclophosphamide as pretransplantation con-
ditioning therapy: a phase I study. Biol Blood Marrow
Transplant 2000; 6: 548–554.
Extramedullary relapse of acute leukemia after allogeneic HCT
K-H Lee et al
842
Bone Marrow Transplantation