High-Dose Chemotherapy and Autologous Bone Marrow Transplantation in Acute
Myeloid Leukemia
ByA.K. McMillan, A.H. Goldstone, D.C. Linch, J.G. Gribben, K.G. Patterson, J.D.M. Richards, I. Franklin,
B.J. Boughton, D.W. Milligan, M. Leyland, R.M. Hutchison, and A.C. Newland
For younger patients with acute myeloid leukemia (AML),
an allogeneic transplant from a matched sibling may afford
the best chance of cure. In patients who are older or
without a matched sibling donor, dose intensification can
be achieved with an autologous bone marrow transplant
(ABMT). We report here the results of a high-dose chemotherapy regime with nonpurged ABMT in 82 adult patients
in first remission of AML with a median follow-up of 31
months. The median age was 40 years (range 16 to 57
years). The median interval between remission and ABMT
was 5 months (range 1 to 12 months). Twenty-eight of
these patients received a second course of the same
high-dose chemotherapy and ABMT. The procedure related mortality rate was 6%. The projected leukemia-free
survival (LFS) at 5 years is 48% for all 82 patients and 50%
for the 76 patients with no known preceding myelodysplastic syndrome. For those patients with primary AML who
received a double ABMT the projected LFS is 67%. The
interval between remission and ABMT did not predict for
either relapse or LFS. ABMT using a multidrug chemotherapy protocol is less toxic than allogeneic BMT yet results in
a similar LFS.
0 1990 by The American Society of Hematology.
C
if only offered to those less than 30 years with a matched
sibling, will be available to less than 10% of patients with
AML.
To extend the role of very intensive consolidation therapy
to a wider patient population, autologous bone marrow
rescue has been used.I3-l7Although there is a risk of leukemic
contamination of the harvested marrow and there will be no
“allogeneic graft-versus-leukemia effect,” the morbidity of
the procedure is likely to be less than with allografts due to
the freedom from problems with either graft-versus-host
disease or ongoing immunosuppression.
We report here that patients with A M L treated in first
remission with high dose chemotherapy and autologous bone
marrow transplantation (ABMT) have a similar projected
survival to patients treated by allogeneic bone marrow
transplantation.
URRENT CONVENTIONAL chemotherapy in adults
with acute myeloid leukemia (AML) results in a high
remission rate with a subsequent high relapse rate. Several
studies have indicated that less than one third of patients
obtaining remission are cured of their disease.’-5 This has
lead to a search for an effective postremission therapy in
AML. The use of maintenance therapy has not been shown to
improve leukemia-free survival (LFS).2*6 Single-center
studies7 and data from registrie~’.~have suggested that
long-term survival is improved in patients treated in first
remission by ablative chemo-radiotherapy and allogeneic
bone marrow transplantation, and this has been confirmed
recently in prospective comparative studies”-” in which
actuarial event-free survivals of approximately 50% have
been reported.
Allogeneic transplantation is generally only applicable to
the younger patient with an HLA-identical sibling. Reports
from the Seattle Marrow Transplant Team suggest that the
good results with allografting are mainly restricted to those
patients under the age of 30 years,7 and yet the majority of
patients with AML are over the age of 50 years. In the UK
Medical Research Council (MRC) AML eighth trial 73% of
patients were less than 60 years but only 34% were less than
40 years, and only 21% were less than 30 years.’ Furthermore, since only about 1 in 4 patients in most Western
societies have a matched sibling, allogeneic transplantation,
From the University College and Middlesex Schools of Medicine:
The London Hospital; Queen Elizabeth Hospital and East Birmingham Hospital, Birmingham; and Leicester Royal Infirmary, Leicester, UK.
Submitted December 7. 1989; accepted April 4. 1990.
A.K.M. and J.G.G. were supported by the Leukaemia Research
Fund of Great Britain.
Address reprint requests to A.H. Goldstone, MD. Department of
Haematology. University College Hospital, Gower St. London WCI
6AU. UK.
The publication costs of thisarticle were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C.section 1784 solely to
indicate this fact.
0 1990 by The American Society of Hematology.
OOO6-4971/90/7603-0027$3.00/0
480
PATIENTS AND METHODS
Patients. Eighty-two adult patients have now been treated by
our collaborative group using an identical protocol. The median
follow-up of this group is 3 1 months with a range between 6 and 110
months. All patients with AML in first complete remission were
entered into the program if (1) they gave informed consent; (2) they
had a good performance status (Karnofsky greater than 70%); (3)
age 16 to 40 years with no HLA-compatible sibling donor or age 41
to 60 years, regardless of donor status. However, it must be noted
that because some patients were referred to the transplant centers
already in remission, an element of patient selection by the primary
physician on other undefined criteria cannot be excluded. There were
38 males and 44 females with a median age of 40 years (range 16 to
57 years).
Four patients were excluded as they were found to be in early
relapse at time of assessment for bone marrow harvest, and three
patients relapsed after the bone marrow harvest while awaiting the
transplant procedure. Six patients had a preceding myelodysplastic
state (MDS), which in all cases had progressed to AML at time of
commencement of induction therapy.
Therapy before ABMT. The patients were all induced with a
combination of daunorubicin and cytosine arabinoside normally in
combination with 6-thioguanine.The earlier patients received courses
of 1 day of daunorubicin at 50 mg/m2 and 5 days of cytosine
arabinoside at 100 mg/m2 every 12 hours as per the UK MRC AML
eighth trial, while the more recent patients received 3 days of
daunorubicin and 10 days of cytosine arabinoside at the same doses
again as in the more recent MRC AML ninth trial. Some patients
Blood, Vol 76, No 3 (August 1). 1990: pp 480-488
ABMT IN ACUTE MYELOID LEUKEMIA
who failed initial induction therapy received second line induction
therapy with other agents such as amsacrine, mitozantrone and
etoposide. The treatment plan was for patients to receive two
consolidation courses with daunorubicin and cytosine arabinoside
after the achievement of a complete remission; however, some
patients received a different number of consolidation courses, as can
he seen in Tables 1 through 3. The mean number of consolidation
courses received was 2.3 with a range from 0 to 6. Sixteen of the 82
patients in this report were previously reported in 1986.13 The
median time to achieve remission for the whole group was 51 days
(range 20 to 338 days), and the median time from attainment of
remission to ABMT was 162 days (range 23 to 365 days). For the
patients without a preceding MDS who received a double graft, the
median time to achieve remission was 40 days (range 20 to 94 days),
and the interval between remission and ABMT was 112 days (range
57 to 277 days). For the patients without a preceding MDS who
received a single graft the time to achieve remission was 61 days
(range 24 to 369 days), and the time from remission to ABMT was
190 days (range 23 to 365 days).
Bone marrow harvest. All patients had a marrow aspirate
performed before harvesting to confirm, on morphologic analysis,
that they remained in remission. Bone marrow was harvested under
general anesthesia and cryopreserved in the manner previously
described." No in vitro purging was undertaken. Patients who did
not have a Hickman line in situ had one inserted at the time of the
harvest. Patients underwent a second bone marrow harvest as soon as
possible after recovery from the first procedure was complete.
Transplunt procedure. The high-dose chemotherapy,which was
identical in all cases, was BCNU 300 mg/m2 day - 5; doxorubicin
50 mg/m2 day -5; cyclophosphamide 1.5 g/m2 day - 5 to -3 with
mesna 1 to 2 g/g of cyclophosphamide; cytarabine 100 mg/m2 every
12 hours day - 5 to -2, and 6-thioguanine 100 mg/m2 every 12
hours day - 5 to -2. The autologousmarrow was thawed rapidly at
the bedside and reinfused on day 0. Patients who achieved satisfactory hematologic regeneration from the first procedure
(neutrophils 20.5 x 109/L and unsupported platelet count of
250 x 109/L by Day 42), and in whom there was no contraindication to further intensive therapy, were advised to proceed to a
second ABMT as soon as possible after the first with a second cycle
of the same high-dose chemotherapy.No posttransplant therapy was
administered while patients remained in remission. All patients were
treated in single rooms with simple reverse barrier nursing procedures without laminar flow facilities. Blood products were not
irradiated and were not screened for cytomegalovirus. Platelets were
given prophylactically when the count was below 20 x 109/L. No
prophylactic antibiotics were administered during the period of
neutropenia, but empirical antibacterial therapy was given for
febrile neutropenic episodes.
Statistical methods. Statistical analyses were performed by
Student's test and chi-squared analysis. Survival curves were drawn
by the Kaplan-Meier life table method and compared by the
log-rank test. The relationship of clinical features to the outcome of
the procedure were analyzed by univariate analysis. Events were
included in the analysis up to March 1, 1990.
RESULTS
Hematologic recovery. For all 82 patients the median
number of days, after the first marrow infusion, to achieve a
total neutrophil count of 20.5 x 109/L was 22 (range 11 to
73), and the median number of days to achieve a platelet
count of 2 5 0 x 109/L, unsupported by platelet transfusion,
was 34 (range 13 to 300). The median number of days after
the second ABMT procedure to achieve a total granulocyte
count of 20.5 x109/L was 27 (range 12 to 48), and the
48 1
median number of days to achieve a platelet count of
2 5 0 x 109/L, unsupported by platelet transfusion, was 38
(range 21 to 98). Paired analysis of the times for hematologic
recovery after the first and second procedures in the 26
patients who received two grafts showed a highly significant
delay for both neutrophil ( P = .004) and platelet recovery
( P = .001) after the second procedure.
Selection of patients for second ABMT. There were four
procedure-related deaths from the first ABMT procedure
(5%), and nine patients relapsed within 90 days of the first
procedure. Five of these nine patients who relapsed early had
temporally recovered hematologically before their early
relapse prevented the second procedure. Of the remaining 69
patients, 54 (78%) achieved full hematologic reconstitution
by day 42. These 54 patients were eligible to proceed to the
planned second ABMT; however, only 25 of these patients
actually received the second cycle of chemotherapy and
ABMT, and in addition three patients who had regenerated
after day 42 following the first procedure (unique patient
numbers [UPNs] EB2, EB4, and L06) also proceeded. The
reasons for cancellation of the second graft in the 29 patients
who did not proceed were: patient refusal (22 patients);
cardiac problems (3 patients: 1 poor left ventricular function,
1 unexplained bradycardia, and 1 unexplained tachycardia);
non-A non-B hepatitis (1); severe pseudomonas cellulitis in
part one (1); failed second bone marrow harvest (1); and
severe psychological problems during part one (1).
Procedure-related morbidity and mortality. All patients
developed fever and required antimicrobial therapy. There
were no cases of pneumonitis during a first procedure. The
causes of the procedure-related deaths that occurred after a
first procedure were cerebral hemorrhage (2 cases); hepatorenal failure ( l ) , and aspergillus pneumonia (1). There were
two cases of pneumonitis during a second procedure (UPN:
UCH 307 and UPN:UCH 323), and both were due to
cytomegalovirus and neither was fatal. However, one (UPN:
UCH 307) had not recovered normal respiratory function by
the time of her death, after relapse, 11 months later. There
was a single death after a second procedure due to a varicella
pneumonia. Overall, in 110 procedures in 82 patients there
have been five procedure-related deaths. Therefore, the risk
of a procedure-related death is 4.5% per procedure or 6%
overall per patient. There were no cases of graft-versus-host
disease despite the fact that random donor platelet and red
blood cell transfusions were not irradiated. There has been no
long-term morbidity from the procedure, except for the one
case referred to previously, and all the survivors have a
Karnovsky status of 100%. There has been a single case of
late pneumonitis (greater than 3 months) that occurred in a
patient who had received a double graft (UPN:UCH245),
but this resolved completely with a short course of oral
steroid therapy. The cause of this pneumonitis was ascribed
to the cumulative dose of BCNU received. There has been no
case of cataract in any patient.
Outcome. The median follow-up for the whole group is
31 months, and a t 5 years the projected overall survival is
53% with a relapse rate of 48% and a projected LFS of 48%
(Fig 1). The outcome in the six patients with a known
preceding myelodysplastic state has been poor, with only 2 of
~~
Sex
M
F
M
M
F
M
820
822
825
83
84
B6
E83
E85
7
25
M
UCH248
M2
DNR, Ara-C, 6TG
2
1
55
M
UCH236
2
39
F
UCH220
M1
DNR, Ara-C, 6TG
4
1
M4
46
F
M4
DNR. Ara-C, 6TG
4
3
UCH 194
DNR. Ara-C, 6TG. m-AMSA, AzaCyt, VP16, cyclo,
2
4
M2
F
UCH 174
M4
F
UCH 163
8
11
3
(Continued on following page)
DNR, Ara-C, Are-C:ID, 6TG. m-AMSA, VP16, mitoz
DNR, Ara-C. 6TG. cyclo, vinc. pred
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
40
36
M
UCHl6O
9
36
19
F
UCH 156
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. vinc. pred
5
4
3
45
M
UCH 132
DNR. Ara-C, 6TG
4
M1
3
M4
45
F
UCH131
DNR, Ara-C, Ara-C:ID, 6TG, m-AMSA, VP16, mitoz
6
3
1
M2
32
F
UCH 128
M2
2
M5
27
F
LO4
DNR, Ara-C, 6TG. m-AMSA. VP16
DNR, Ara-C, 6TG. m-AMSA, VP16
7
6
2
3
M4
20
M
LO3
M1
4
M4
42
F
LO2
TD
2
2
Hepatitis
Cardiac
Slow recov
2
3
Relapse
4
Cardiac
Cardiac
0
1
Refused
Failed 2nd BMH
Slow recov
Slow recov
Relapsed
Refused
Refused
5
1
6
2
3
2
2
Refused
Slow recov
Slow recov
1
DNR, Ara-C, 6TG
Refused
Slow recov
2
1
DNR, Ara-C, 6TG
DNR, Ara-C. 6TG
2
2
DNR, Ara-C, 6TG
Relapsed
3
2
DNR, Ara-C, 6TG
Refused
Slow recov
4
Refused
Relapsed
Refused
Refused
Slow recov
Refused
Refused
TD
Refused
Reason for
No Part 2
DNR, Ara-C, 6TG. cyclo, vinc. pred
DNR, Ara-C, 6TG
DNR. Ara-C, 6TG. m-AMSA, VP16
2
3
0
DNR, Ara-C,6TG
1
2
DNR, Ara-C, Ara-CID, 6TG. mitoz
DNR. Ara-C, 6TG
3
3
5
5
4
4
No. of
Consolidation
Courses
DNR, Ara-C, 6TG. cyclo, vinc. pred
DNR, Ara-C, 6TG. m-AMSA, AzaCyt. VP16
DNR, Ara-C, 6TG. cyclo, vinc, pred
11
2
28
M4
40
M
F
LE2
LO 1
M4
M3
32
F
LE 1
4
6
1
3
6
M3
39
7
10
6
10
1
6
2
1
2
2
3
1
6
11
1
3
12
DNR, Ara-C, 6TG. cyclo, vinc. pred
DNR, Ara-C, 6TG. m-AMSA, AzaCyt, VP16
8
DNR. Ara-C. 6TG. cyclo, vinc. pred
Drugs in InductionIConsolidation
10
8
CR to
ABMT
(mos)
2
2
2
2
(mos)
Diagnosis
to CR
2
M2
28
F
M
E88
M2
23
F
E86
M3
43
M
M2
55
M4
M4
42
49
M5
M4
M4
M2
M1
M4
M4
FAB
TvDe
43
24
48
44
F
819
51
57
M
M
B16
54
812
M
~
81 1
UPN
Ageat
T x ( d
Table 1. Single-Autograft Data
~~
41
6
2
4
0
7
65
7
11
73
74
3
9
8
18
4
55
20
26
28
2
58
6
2
12
9
25
26
26
1
48
LFS (mod
Status
Alive-cr
DR
TO
DR
DR
OR
Alive-cr
DR
DR
Alive-cr
Alive-cr
DR
AR
DR
AR
DR
Alive-cr
OR
Alive-cr
Alive-cr
DR
Alive-cr
OR
DR
OR
AR
Alive-cr
Alive-cr
DR
TD
Alive-cr
P
P
--I
m
N
Q)
21
M
828
M2
M4
M4
1
1
1
2
12
1
1
1
1
1
1
3
2
1
1
1
1
1
1
Diagnosis
to CR
lmos)
DNR, DOX, Ara-C, Ara-C:ID, BTG, VP16. mitoz
7
10
10
6
5
DNR, Ara-C, 6TG
DNR, Ara-C. 6TG. AraC:ID. m-AMSA, VP16, mitoz
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, AraC, 6TG. AraC:ID, m-AMSA, VP16, mitoz
DNR, AraC, 6TG. AraC:ID. m-AMSA, VP16. mitoz
DNR, Ara-C, 6TG
6
DNR, AreC, 6TG
3
DNR, AraC, 6TG
DNR, AraC, 6TG
DNR, AraC, 6TG
DNR, Ara-C, 6TG
Relapse
3
4
2
2
0
3
Slow recov
Refused
Refused
7
10
8
14
9
Slow recov
Refused
10
14
10
11
2
22
1
1
27
11
103
32
1
1
LFS lmos)
Slow recov
Refused
Refused
2
Refused
Relapsed
Slow recov
TD
2
2
2
2
2
1
Psych problem
2
DNR, Ara-C, 6TG. cyclo. vinc. pred
Slow recov
2
2
Inf in part 1
Slow recov
Relapse
TD
Reason for
No Part 2
DNR, AraC, 6TG
2
2
2
No. of
Consalidation
Courses
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C. 6TG
Drugs in Induction/Consolidation
3
10
5
3
4
2
8
4
7
4
5
6
CR to
ABMT
Imos)
Alive-cr
Alive-cr
AR
Alive-cr
Alive-cr
Alive-cr
Alive-cr
AR
Alive-cr
DR
Alive-cr
TD
DR
Alive-cr
Lost t o follow-up
Alive-cr
Alive-cr
DR
TD
Status
Abbreviations: Tx. time of transplantation; Diagnosis t o CR. time interval between diagnosis and remission; CR to ABMT, time interval between remission and first transplant: Drugs in Induction/Consolidation, chemotherapy drugs
received during induction and consolidation; Reason for No Part 2. reason for not proceeding t o second graft: DNR, daunorubicin; Ara-C, cytarabine: 6TG. 6-thioguanine; cyclo, cyclophosphamide, vinc. vincristine: pred, prednisalone;
m-AMSA. amsacrine; AzaCyt. azacytidine; VP16. etoposide: Ara-C:ID, intermediate-dose cytarabine; mitoz, mitoxantrone; TD, toxic death; Slow recov. slow hematologic reconstitution; Failed 2nd BMH. failed bone marrow harvest:
Cardiac, cardiac problem preventing second part: Inf in part 1, serious infection in part 1: Psych problem. psychologicalproblem preventingsecond part; Alive-cr, alive in remission; DR, dead in relapse: AR, alive in relapse; DOX, doxorubicin.
48
M
827
M3
24
41
F
M
829
M3
24
F
LO6
E26
M1
51
F
E812
M4
18
M2
30
M
F
M2
E810
38
F
E81 1
M2
M2
M2
M1
M5
M4
M3
M3
M4
M6
FAB
Type
E89
44
34
F
F
UCH339
UCH356
48
49
F
M
45
M
UCH311
UCH328
34
UCH325
38
M
F
50
M
UCH287
UCH30
54
UCH303
48
F
F
UCH264
Sex
UPN
UCH267
Age at
Tx(~s)
Table 1. Single-Autograft Data (Cont’d)
P
w
OD
a5
x
r
m
C
P
0
2
r
3
;I
C
b
z
5-
484
McMlLLAN ET AL
Table 2. Double-Autograft Patient Data
CR to
ABMT
Ageat
Type
Diagnosis
toCR
UPN
Sex
Transplant
lpl
(mosl
(mod
UCH 19
UCH 105
UCH108
UCH 143
UCH 147
UCH 148
UCH209
UCH2 18
M
F
F
M
M
M
M
M
24
44
39
48
35
26
17
26
M1
M4
M1
M1
M5
M1
M2
M4
2
2
1
1
2
1
1
1
2
6
8
3
4
5
4
7
UCH223
UCH237
UCH241
UCH245
F
M
F
M
43
16
55
46
M4
M1
M5
M4
1
3
2
1
2
3
3
5
UCH259
M
16
M2
2
4
UCH272
UCH307
UCH320
F
F
F
32
21
41
M2
M5
M1
1
1
1
3
5
3
UCH324
UCH33 1
UCH38 1
M
F
M
26
48
45
M4
M2
M4
1
1
2
5
3
6
E10
M
23
M4
2
9
E8 1
E82
EB4
EB7
F
M
F
30
34
47
37
M1
M5
M3
M2
1
1
2
3
3
3
3
6
LE3
UCH 426
F
F
36
35
M2
M4
1
1
9
3
FAB
F
Time
Between
Grafts
LFS
lmos)
(mosl
2
1
4
2
3
2
2
3
3
2
2
110
80
79
69
69
68
49
21
Alive-cr
Alive-cr
Alive-cr
Alive-cr
Alive-cr
Alive-cr
Alive-cr
AR
2
1
2
5
2
4
4
5
47
44
15
42
Alive-cr
Alive-cr
DR
Alive-cr
2
2
18
2
2
3
3
4
2
36
9
25
Lost to
follow-up
Alive-cr
DR
Alive-cr
2
2
2
3
2
3
125
10
16
5
4
7
2
2
2
2
3
3
2
3
6
20
39
14
TD
DR
Alive-cr
AR
1
2
2
3
38
8
Alive-cr
Alive-cr
No. of
Drugs in Induction/
Consolidation
Consolidation
Courses
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. cyclo,
vinc, pred
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. VP 16
DNR, Ara-C, 6TG. VP16
DNR, Ara-C, 6TG. cyclo,
vinc. pred
DNR, Ara-C. 6TG
1
2
6
1
3
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Are-C, 6TG. cyclo,
vinc. pred
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, ID ARA-C,
mAMSA, VP 16
DNR, Ara-C, 6TG. cyclo,
vinc. pred
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. cyclo,
vinc, pred
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
Status
Alive-cr
DR
Alive-cr
DR
Abbreviations: Time between grafts, interval between first and second BMT; see Table 1 for other abbreviations.
6 remaining in remission at 12 months and 28 months
post-ABMT. In the 76 patients with an apparently primary
disorder at diagnosis, the projected overall survival at 5 years
is 55%, with a relapse rate of 47% and a projected LFS of
50% (Fig 2). There have been no relapses later than 26
months after ABMT. Of the 27 patients who relapsed there
are only two survivors of greater than 1 year’s duration after
relapse: one following a second remission busulphan and
cyclophosphamide autograft (UPN:UCH 21 8) and a second
who relapsed with a myelodysplastic marrow picture (UPN:
UCH 236). In the 26 patients with an apparently primary
disorder who completed the double autograft protocol, the
projected overall survival at 3 years was 73%, with a
projected relapse rate of 30%and projected LFS of 67%.
To analyze the effect of the second graft, the survival of 25
patients with an apparently primary disorder who had a
Table 3. MDS Patient Data
Time
Age at
UPN
Sex
~~
Transplant
FAB
Ivrsl
Tvoe
~
~
Diagnosis
toCR
(mas)
CR to
ABMT
(mod
M
M4
M2
M4
M2
M6
1
4
1
2
4
4
8
6
3
9
F
34
M1
11
4
F
M
F
LO5
F
Reason
No.
Drugs in Induction/
Consolidation
Consolidated
Courses
fwNo
Part2
of
Grafts
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. VP16
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG
DNR, Ara-C, 6TG. m-AMSA,
AzaCyt, VP 16
DNR, Ara-C, , Ara-C:ID, 6TG.
m-AMSA, VP 16, mitoz
2
3
2
2
4
Relapse
Refused
Relapse
-
Refused
1
1
1
2
1
0
-
2
Between
Grafts
(mo4
LFS
(mod
Status
-
4
12
3
7
28
DR
DR
DR
DR
Alive-cr
4
12
Alive-cr
~~
55
55
50
39
45
UCH215
UCH231
UCH239
UCH323
E17
No.of
Abbreviations: see Tables 1 and 2.
-
3
485
ABMT IN ACUTE MYELOID LEUKEMIA
80
li
Leukaemia
Free
Survival
("A)
_1
20
J
Fig 1.
2
3
4
5
b
'I
8
4
TIME (YEARS)
remission, interval from remission to transplant, and the
number of consolidation courses given. No factor was significant in predicting relapse or LFS. In particular, when the
patients are divided into two groups, those grafted within 6
months of remission (n = 50) and those grafted between 6
and 12 months into remission (n = 32), there is no significant difference between the two groups in terms of LFS (49%
v 46%). If those who only received a single ABMT are
analyzed alone there is a small advantage (P = .14) to those
receiving ABMT between 6 months and 1 year into remission, but this does not reach significance.
single graft with rapid hematologic recovery and remained in
remission at 90 days was examined. The median interval
between grafts in the double-graft group was 90 days. Rapid
hematologic recovery was defined as up to 49 days, as this
was the latest day on which a patient with primary AML
recovered from the first graft and then received a second
procedure. However, this comparison single-graft group
differed from the double group in that the interval between
remission and ABMT was significantly longer (P = .003).
The projected overall survival at 3 years for this group who
did not receive the second graft for a nonhematologic reason
was 47%, with a projected relapse rate of 60% and a
projected LFS of 40%. The difference between these patients
who did not receive the second graft for a nonhematologic
reason and those who did receive the second graft approaches
significance (P = .052) (Fig 3). Seven patients receiving a
double ABMT for primary AML have relapsed, and four of
these relapses occurred beyond 1 year from the first ABMT,
while only 2 of 14 relapses occurring in the comparison group
occurred after 1 year.
Prognostic factors for outcome of ABMT that have been
analyzed by univariate analysis included age, sex, FrenchAmerican-British (FAB) type, interval from diagnosis to
Leukaemia
Free
Survival
'1
LFS of all 82 patients.
DISCUSSION
The chemotherapy autografting procedure used in this
study has resulted in a low early procedure-related mortality
(6%) compared with the 20% procedure-related deaths in the
first 100 days after allogeneic BMT reported by the Seattle
group.' In addition, in allograft recipients there is also a 10%
incidence of late procedure-related deaths7compared with no
late nonleukemic deaths after ABMT in this study. There
has been almost no significant long-term morbidity from
cataracts or pneumonitis observed with this regime, and this
contrasts with the sequelae of conventional allografting
1h
80
("w
"PRIMARY" AML patients n =76
MDS patients n = 6
p is NS.
'1
2
3
4
5
6
TIME (YEARS)
'I
8
4
Fig 2. LFS of 76 patients without evidence of a
preceding myelodysplastic syndrome.
486
Leukaemia
Free
Survival
(W
McMlLLAN ET AL
I0rk4
80
I
I
Lu
“‘“It 1-i
2o
t1
I
I I I I I
Ill
I
I1
0
p = 0.052
‘1
2
3
4
5
6
TIME (YEARS)
protocols. We believe that the early procedure-related mortality and late morbidity are also reduced when compared with
autografting regimes that use total body irradiation (TBI).
Comparisons with autografts using busulphan and cyclophosphamide without TBI are difficult because this has been used
mainly in second remission; although in that context it has
considerable toxicity, it is not possible to say if this will also
be true of its use in first remission. However, the early
morbidity with our protocol is considerable compared with
conventional therapy as evidenced by the fact that only 41%
of patients alive in remission after the first procedure
proceeded to the second autograft as originally planned.
The rationale of the double-autograft program is based on
two premises. Reharvesting of the bone marrow after completion of the first course of chemotherapy when leukemic
contamination is likely to be at a nadir effects an in vivo
purge, and the two-stage process allows the overall dose of
drug that is delivered to be increased without a corresponding increase in toxicity. For three reasons there has been
failure to complete the double-graft program in the majority
of patients. First, there is a population of patients who
relapse early (less than 90 days) after the first graft and who
cannot therefore receive the second graft. Second, as regeneration after the second course of intensive chemotherapy is
significantly delayed compared with the first, in those cases
where recovery from part 1 was slow it was thought unjustified to proceed to a second graft. This selection has kept the
procedure-related mortality of the second graft to a minimum but reduced the proportion receiving it. Third, because
the first autograft follows intensive induction and consolidation therapy, some patients are unwilling to complete such an
arduous treatment program.
The advantage of decreased toxicity compared with allogeneic BMT is balanced by the increased relapse rate seen after
ABMT (48% v 25% in the Seattle study of allogeneic BMT in
first remission7). However, the projected LFS at 5 years for
the two groups is very similar, with an LFS of 48% in this
study compared with 45% in a recent report from Seattle”
and 48% reported by the International BMT Registry’ for
allogeneic BMT in AML in first remission. It is, we believe,
7
8
4
Fig 3. LFS: double ABMT versus single ABMT
in patients with hematologic recovery by day 49
who remain in remissionat 90 days.
justifiable to compare these two transplant groups as it is
likely that any selection in the two groups will be similar. The
favorable comparison of these results with those of allogeneic
BMT is underlined by the fact that the age of many of these
patients would have put them in a high-risk group for
transplant-related complications after allogeneic BMT.’,19
They can also be compared favorably with the results of
autografting regimes using TBI reported to the European
Bone Marrow Transplant Group (EBMT)?’where the LFS
at 5 years is 42%. A study of 22 patients from Germany using
TBI has reported an LFS of 6195.’’ There is an apparent
advantage in favor of the use of a double-autograft procedure
when patients who received a second procedure are compared
with those who did not receive a second graft for a nonhematologic reason, although this does not quite achieve statistical
significance (P = .052). Furthermore, it is noteworthy that
this group has a significantly shorter remission to transplantation interval than the comparison group. The only definitive way of resolving the value of the second graft would be to
randomize patients who are ready to proceed, immediately
before the second procedure, to proceed to a second graft or
not. This would pose major logistical difficulties.
No in vitro purging procedures, using cyclophosphamide
derivatives, were used in this study as we believe that, unlike
the situation in the rat model,2’ there is no clear in vitro
evidence of their selectivity for leukemic cells in human^.^^.*^
The use of these drugs may slow recovery and therefore
increase the risk of procedure-related death. Recovery times
were observed in this study that were shorter than those
reported with regimes using TBI, especially if combined with
3
hematologic recovery was
purged marrow. 1 4 ~ ’ 5 ~ 1 7 . 2Although
slower with the second autograft, no patient in remission
remained platelet-dependent beyond 100 days, which is not
infrequent with purged marrow.14si5*17,23
The most widely
quoted evidence that purging may be efficacious in AML
autografting comes from retrospective analysis of the EBMT
registry,20which suggests that in a group of patients ablated
with cyclophosphamide and TBI, LFS was significantly
greater in those patients whose marrow was purged with
derivatives of cyclophosphamide and that this advantage was
487
ABMT IN ACUTE MYELOID LEUKEMIA
most marked in patients transplanted less than 6 months into
remission. This is multicenter data in which some centers
performed purging and others did not, and it may be
susceptible to differing selection policies. It must also be
noted that the excellent results reported by Korbling et all’
were obtained with purged marrow.
The key question in this study is whether survival has been
improved compared with using conventional chemotherapy
alone. Unlike the comparison of transplant series, the effect
of selection of patients for transplantation and the effect of
the delay between remission and transplantation (“time
censoring”) needs to be considered when these results are
compared with those of conventional chemotherapy. As this
group of patients had been in remission for a median of 5
months before ABMT we have compared their outcome to
that of patients in the ninth MRC A M L trial who maintained first remission for a minimum of 6 months. This shows
that compared with the projected LFS of 48% a t 5 years
reported here, the projected LFS a t 4 years with conventional
therapy in the ninth MRC A M L trial of those in remission
for a minimum of 6 months is 40%. (R. Gray, personal
communication, November 1989). The difference is clearly
small and any advantage for ABMT is attributable to those
who received the double procedure. However, comparison of
patients in different studies cannot justify a definite conclu-
sion, and the results of large prospective randomized trials
are needed. A recent small trial comparing allogeneic BMT,
ABMT, and conventional chemotherapy reported a significant advantage to allogeneic BMT compared with conventional chemotherapy.” The results of autografting were
intermediate and not significantly different from either the
allogeneic BMT or the conventional therapy groups. Confirmation of these findings will have to await the results of
larger studies. When the results of comparative trials are
examined it must be remembered that the ablative therapy
used may influence the outcome; eg, the use of single-agent
high-dose melphalan as opposed to TBI or a multidrug
regime in the study by Reiffers et allz may have affected the
outcome of the ABMT arm.
In conclusion, we have found that ABMT using a multidrug chemotherapy protocol appears less toxic than allogeneic BMT, yet results in a similar LFS. The relatively low
mortality and morbidity is particularly important when the
use of autografting is extended, as in this study, to patients
up to age 60 years, which thus enables intensive therapy to be
applied to a much higher proportion of the A M L patient
population than allogeneic BMT. The results of ABMT
compare favorably with conventional consolidation and maintenance therapy, but definitive proof of improved survival
must await the outcome of prospective randomized trials.
REFERENCES
1. Rees JHK, Gray RG, Swirsky D, Hayhoe FGJ: Principal
results of the Medical Research Council’s 8th Leukaemia Trial.
Lancet Vol2:1236, 1986
2. Preisler H, Davies RB, Kirshner J, Dupre E, Richards F,
Hoagland C, Kopel S, Levy R, Carey R, Schulman P, Gottleib AJ,
McIntyre OR, Cancer and Leukaemia Group B: Comparison of
three remission induction regimes and two post induction strategies
for the treatment of acute non-lymphoblastic leukaemia: A cancer
and leukaemia group B study. Blood 69:1441, 1987
3. Buchner T, Urbanitz D, Hiddemann W, Ruhl H, Ludwig WD,
Fischer J, Aul HC, Vaupel HA, Kuse R, Zeile G, Nowrousian MR,
Konig HJ, Walter M, Wendt FC, Sodomann H, Hossfeld DK, von
Paleske A, Loffler H, Gassman W, Hellriegel P, Fulle HH, Lunscken c h , Emmereich B, Pralle H, Pees HW, Pfeundschuh M,
Bartels H, Koeppen KM, Schwerdtfeger R, Donhuijsen-Ant R,
Maimer K, Bonfert B, Koppler H, Zurborn KH, Ranft K, Thiel E,
Heinecke A: Intensified consolidation and induction with or without
maintenance chemotherapy for acute myeloid leukaemia (AML):
Two multicentre studies of the German AML cooperative group. J
Clin Oncol3:1583, 1985
4. Tallman MS, Kopecky KJ, Amos D, Dahlberg S, Hewlett JS,
Files JC, Coltman CC, Thomas ED, Appelbaum FR: Analysis of
prognostic factors for the outcome of marrow transplantation or
further chemotherapy for patients with acute non lymphocytic
leukaemia in first remission. J Clin Oncol7:326, 1989
5 . Rohatiner AZS, Gregory WM, Bassan R, Barnet MJ, Waxman J, Richards MA, Ganesan TS, Tucker J, Malik TSA, Birds R,
Amess J, Berney S, Malpas JS, Lister T A Short-term therapy for
acute myelogenous leukaemia. J Clin Oncol6:218,1988
6. Rees JHK: The chemotherapy of acute myeloid leukaemia in
UK: Past, present and future. Bone Marrow Transplant 4:110, 1989
(abstr)
7. Clift RA, Buckner CD, Thomas ED, Kopecky KJ, Appelbaum
FR, Tallman M, Stord R, Sanders J, Sullivan K, Banaji M, Beatty
PS, Bensinger W, Cheever M, Deeg J, Doney K, Fefer A, Greenberg
P, Hansen J, Hackman R, Hill R, Martin P, Meyers J, McGuffin R,
Neiman P, Sale G, Shulman H, Singer J, Stewart P, Weiden P,
Witherspoon R: The treatment of acute nonlymphoblastic leukaemia by allogeneic transplantation. Bone Marrow Transplant
2243, 1987
8. Advisory Committeeof the International Bone Marrow Transplant Registry: Report from the International Bone Marrow Transplant Registry. Bone Marrow Transplant 4:221, 1989
9. Gratwohl A, Hermans J, Lyklema J, Zwaan F: Bone marrow
transplantation for leukaemia in Europe. Bone Marrow Transplant
4:1,1989 (abstr)
10. Appelbaum FR, Dahlberg S, Thomas ED, Buckner CD,
Cheever MA, Clift RA, Crowley J, Deeg HJ, Fefer A, Greenberg
PD, Kadin M, Smith W, Stewart P, Sullivan P, Storb R, Weiden P
Bone marrow transplantation after remission induction for adults
with acute non-lymphoblastic leukaemia. Ann Intern Med 101:581,
1984
11. Appelbaum FR, Fisher LD, Thomas ED: Chemotherapy
versus marrow transplantationfor adults with acute non-lymphoblastic leukaemia:A five year follow up. Blood 72:179, 1988
12. Reiffers J, Gaspard MH, Maraninchi D, Michallet M, Marit
G, Stoppa AM, Corront B, David B, Sotto JJ, Broustet A, Carcassome Y,Hollard D: Comparison of allogeneic or autologous bone
marrow transplantation and chemotherapy in patients with acute
myeloid leukaemia in first remission: A prospective controlled trial.
Br J Haematol72:57,1989
13. Goldstone AH, Anderson CC, Linch DC, Franklin IM,
Boughton BJ, Cawley JC, Richards JDM: Autologous bone marrow
transplantation following high dose chemotherapyfor the treatment
of adult patients with acute myeloid Leukaemia. Br J Haematol
64529, 1986
14. Gorin NC, Laporte JP, Douay L, Lopez M, Najman A:
Autologous bone marrow transplantation for consolidation of acute
488
leukaemia in remission: Purging or not purging? Influence of pre
transplant intervals, in KA Dicke, G Spitzer, S Jagannath, EvingerHodges MJ (eds): Proceedings of the IVth International Symposium
on Autologous Bone Marrow Transplantation. The University of
Texas, MD Anderson Cancer Center, 1988
15. Carella AM, Gaozza E, Santini G, Martinengo M, Pungolino
E, Piatti G, Congiu A, Nati S, Carlier P, Giordano D, Cerri R,
Risson M, Miceli S, Raffo MR, Scarpati D, Corvo R, Franzone P,
Vitale V, Vimercati AR, Amico TD, Marmont AM: Autologous
unpurged bone marrow transplantation for acute non-lymphoblastic
leukaemia. Bone Marrow Transplant 3537, 1988
16. Yeager AM, Kaizer H, Santos GW, Sarai R, Colvin OM,
Stuart RK, Braine HG, Burke PJ, Ambinder RF, Barns WH, Fuller
DJ, Davis JM, Karp JE, May WS, Rowley SD, Sensenbrenner LL,
Vogelsang GB, Wingard JR: Autologous bone marrow transplantation in patients with acute non-lymphocytic leukaemia using ex-vivo
marrow treatment with 4-hydroxycyclophosphamide. N Engl J Med
315:141, 1986
17. Korbling M, Hunstein W, Fliedner TM, Cayeux S, Dorken B,
Fehrentz D, Haas R, Ho AD, Keilholz U, Knauf W, Konig A,
Mende U, Pezzutto A, Von Reumont J, Wolf G, Wannenmacher M,
Winkel K, Rother K Disease-free survival after autologous bone
marrow transplantation in patients with acute myelogenous Leukaemia. Blood 74:1898, 1989
McMlLLAN ET AL
18. Linch DC, Knott LJ, Patterson KG, Cowan DA, Harper PG:
Bone marrow processing and cryopreservation. J Clin Pathol35:186,
1982
19. Tallman MS, Kopecky KJ, Amos D, Dahlberg S, Hewlett JS,
Files JC, Coltman CA, Thomas ED, Appelbaum FC: Analysis of
prognostic factors for the outcome of marrow transplantation or
further chemotherapy for patients with acute non-lymphocytic
leukaemia in first remission. J Clin Oncol73:326, 1989
20. Gorin NC, Aegerter P, Auvert B: Autologous bone marrow
transplantation for acute leukaemia in remission: An analysis on
1322 cases. Bone Marrow Transplant 4:3,1989 (abstr)
21. Sharkis SJ, Santos GW, Colvin M: Elimination of acute
myelogenous leukaemic cells from marrow and tumor suspensions in
the rat with 4-hydroxycyclophosphamide. Blood 55:521,1980
22. Singer CRJ, Linch DC: Comparison of the sensitivity of
normal and myeloid leukaemia progenitors to in vitro incubation
with cytotoxic drugs: A study of pharmacological purging. Leuk Res
11:953, 1987
23. Gorin NC, Douay L, Laporte JP, Lopez M, Mary JY,
Nayman A, Salmon C, Aegerter P, Stachowiak J, David R, Pene F,
Kantor G, Deloux J, Duhamel E, Van den Akker J, Gerota J, Parlier
Y, Duhamel G: Autologous bone marrow transplantation using
marrow incubated with Asta Z 7557 in adult acute leukaemia. Blood
67:1367, 1986