Leukemia Research 25 (2001) 379–384

Allogenic stem cell transplantation as salvage therapy for patientsrelapsing after autologous transplantation: experience from a single

institution

Carmen Martınez, Enric Carreras *, Montserrat Rovira, Alvaro Urbano-Ispizua,Jordi Esteve, Francesc Fernandez-Aviles, Marıa Perales, Susana Rives, Marta Gomez,

Emilio MontserratTransplantation Unit, Department of Hematology, Postgraduate School of Hematology ‘Farreras-Valentı’,

Institute of Hematology and Oncology, IDIBAPS Hospital Clınic, Uni6ersity of Barcelona, Villarroel 170, 08036 Barcelona, Spain

Received 4 July 2000; accepted 20 October 2000

Abstract

The prognosis of patients relapsing after an autologous transplant (autoSCT) is very poor. Allogenic stem cell transplantation(alloSCT) offers the possibility of curing some of these patients, at the cost, however, of a high transplant related mortality(TRM). The aim of this study was to analyze the outcome of 14 consecutive patients with hematologic malignancies, from a singleinstitution, who underwent alloSCT for progressive disease after autoSCT. Patients had relapsed at a median of 11.5 months(range 2–72) after autoSCT and they underwent alloSCT at a median of 25.5 months (range 7–73) from the first transplant. Tenpatients received HLA-identical related peripheral blood progenitor cells, three patients underwent matched-unrelated donormarrow transplants, and one patient received a mismatched related transplant. Conditioning regimens consisted of total bodyirradiation plus cyclophosphamide (n=5) or melphalan (n=1), or high-dose combination chemotherapy (n=8). Cyclosporin Aand methothrexate were administered as graft-versus-host disease (GVHD) prophylaxis. Eight patients (57%) developed gradeII–IV acute GVHD. All evaluable patients (n=6) presented extensive chronic GVHD. Overall survival at 1 year was 16%(median 3.5 months, 95% CI 0.7–10.3). Ten patients (71%) died from transplant related complications at a median of 3.5 months(range 0.7–11). Only one patient died of recurrent disease. Three patients remain alive and in complete remission at the time ofthis report (4, 20 and 20 months, respectively). In conclusion, alloSCT offers the possibility of a sustained control of the diseasein some patients who relapse after an autoSCT. However, the procedure is associated with a high transplant-related mortality.Better results might be obtained by carefully selecting patients and by reducing the intensity of the preparative regimen. © 2001Elsevier Science Ltd. All rights reserved.

Keywords: Autologous transplantation; Relapse; Allogenic transplantation

www.elsevier.com/locate/leukres

1. Introduction

Hematopoietic stem cell transplants are effective in avariety of hematologic malignancies [1–3]. Relapse ofthe underlying disease after transplantation remains themost frequent cause of treatment failure, particularlywith autologous transplants (autoSCT), and is generally

considered to have a poor prognosis. The optimaltreatment strategy for these patients is not resolved [4].Although in some patients a second complete remissioncan be obtained with conventional chemotherapy, veryfew patients, if any, are cured with this approach. Therole of second transplant in the treatment strategy ofpatients relapsing after a first transplant is controver-sial, basically because of the treatment-related toxicity.

Allogenic stem cell transplantation (alloSCT) hasbeen employed as a salvage therapy in patients whorelapse after a first alloSCT. The results from largeseries of leukemia patients reported by the both Eu-

Abbre6iations: AlloSCT, allogenic stem cell transplantation; Au-toSCT, autologous transplant; GVHD, graft-versus-host disease;TRM, transplant related mortality.

* Corresponding author. fax: +34-93-2275428.E-mail address: carreras@clinic.ub.es (E. Carreras).

0145-2126/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.PII: S 0 1 4 5 -2126 (00 )00146 -6

C. Martınez et al. / Leukemia Research 25 (2001) 379–384380

ropean Group for Blood and Marrow Transplantationand by the International Bone Marrow Transplantationregistries have shown that a minority of the patientsachieve long-term disease-free survival. However, a sub-stantial proportion of the patients die due to the toxic-ity related to the procedure or because of the relapse ofthe disease [5,6]. Treatment results are better in youngpatients, individuals with a long time interval betweenthe first and the second transplant, and those with goodperformance status.

Experience with alloSCT in patients who relapseafter autoSCT is limited, and factors that could predicta favorable outcome in this group of patients have notbeen fully established [7–10].

The purpose of this study was to evaluate the out-come of patients with hematologic malignancies whounderwent alloSCT for progressive disease after auto-SCT in our institution.

2. Patients and methods

2.1. Patient population

Fourteen alloSCT were performed between March1995 and December 1999 for the treatment of hemato-logic malignancies relapsing after autoSCT. Five pa-tients had non-Hodgkin’s lymphoma, four acuteleukemia, three Hodgkin’s disease, one multiplemyeloma, and one chronic myeloid leukemia. The out-come of two patients with mantle-cell lymphoma allo-grafted after relapsing from an autoSCT has beendescribed in a separate publication [11]. The clinicalcharacteristics of these patients prior to both the firstand second transplants are described in Table 1. Themedian age of patients at the time of alloSCT was 37years (range 20–57). Nine patients were male and fivewere female. The median time interval from diagnosisto autoSCT was 15 months (range 4–44). Patients hadrelapsed at a median of 11.5 months (range 2–72) afterautoSCT. All patients but one were treated with one(n=10) or ] two (n=3) chemotherapy regimens be-fore alloSCT. Only three patients were in completeremission (CR) at the time of the second transplant,and one patient with CML was maintained in thechronic phase. AlloSCT was performed at a median of25.5 months (range 7–73) following autoSCT. Themedian Karnofsky performance status score prior toalloSCT was 70% (range 40–80%).

2.2. Transplantation methods

The preparative regimen for autoSCT consisted ofcyclophosphamide (Cy) 120 mg/kg and fractionatedtotal body irradiation (TBI) 12 Gy in four patients,melphalan 200 mg/m2 in one, and high-dose combina-

tion chemotherapy in nine (Table 1). The stem cellsource for the first transplant was bone marrow (n=3),peripheral blood (n=10), or both (n=1).

The conditioning regimen for alloSCT consisted ofCy 120 mg/kg and TBI 12–13 Gy in six cases, busulfan(Bu) 16 mg/kg and Cy 120 mg/kg in four, melphalan140 mg/m2 and TBI 12 Gy in one, Cy 120 mg/kg+Bu16 mg/kg+etoposide 30 mg/kg in one, and fludarabine150 mg/m2+melphalan 140 mg/m2 in two. Ten pa-tients received HLA-identical related (sibling (n=9),father (n=1)) peripheral blood progenitor cells, threeunderwent matched-unrelated donor marrow trans-plants, and one patient received a mismatched relatedperipheral blood transplant.

All patients received cyclosporin A and short coursemethothrexate as graft-versus-host disease (GVHD)prophylaxis. The diagnosis and grading of acute andchronic GVHD was established according to the Seattlecriteria [12]. Chronic GVHD was defined as the pres-ence of GVHD after day 90.

2.3. Statistical methods

Median values and ranges for age and time intervalswere determined. Survival was measured from the dateof alloSCT to the date of death from any cause, or lastknown follow-up. The method of Kaplan and Meierwas used to plot survival and relapse curves. Theanalysis was performed with SPSS software.

3. Results

3.1. Hematopoietic engraftment

One patient died before hematopoietic engraftment.The median time to achieve an absolute neutrophilcount of \0.5×109/l and \1.0×109/l was 15 days(range 10–23), and 17 days (range 11–25), respectively.The median time to achieve a platelet count \20×109/l was 15 days (7–38) days; five patients did notachieve this platelet count after transplant. No patientrejected the graft.

3.2. GHVD

Eight patients (57%) developed acute GVHD gradeII–IV. Six patients survived more than 90 days andwere evaluable for chronic GVHD. All of them pre-sented extensive chronic GVHD; this complication wasfatal in two cases (patients c3 and c4). Two patients(cases c2 and c7) died due to infectious complica-tions associated with immunosuppressive therapy forchronic GVHD (Table 2).

C. Martınez et al. / Leukemia Research 25 (2001) 379–384 381

Tab

le1

Pat

ient

char

acte

rist

icsa

Dia

gnos

isT

ime

toA

utoS

CT

Tim

eto

Allo

SCT

Tim

ebe

twee

nSe

xP

atie

ntA

geat

allo

SCT

auto

SCT

–re

laps

ebau

toSC

Tb

Stem

cell

Con

diti

onin

gC

ondi

tion

ing

Stem

cell

sour

ceal

loSC

Tb

Kar

nofs

kyP

Sso

urce

regi

men

regi

men

01C

yTB

IM

7273

HL

A-m

atch

edU

NR

BM

BuC

Y80

26A

ML

11B

MB

uCy

1239

HL

A-m

atch

edU

NR

BM

CyT

BI

BM

80C

ML

3102

M48

BM

+P

BF

BuC

y9

12H

LA

-mis

mat

ched

sibl

ing

PB

CyT

BI

6027

AM

L10

03P

BF

BE

AC

1026

HL

A-m

atch

edsi

blin

gP

BC

yTB

I80

38N

HL

4404

ML

F9

19H

LA

-mat

ched

sibl

ing

PB

ML

FT

BI

PB

2005

4M

M48

M23

PB

BE

AC

1129

HL

A-m

atch

edsi

blin

gP

BB

uCyV

P50

0636

MH

DB

EA

C24

33H

LA

-mat

ched

sibl

ing

PB

CyT

BI

PB

60F

0735

NH

L45

PB

FB

EA

C24

30H

LA

-mat

ched

sibl

ing

PB

CyT

BI

8053

MC

L15

08P

BM

BE

AC

1319

HL

A-m

atch

edsi

blin

gP

BC

yTB

I80

57M

CL

1509

CyT

BI

711

HL

A-m

atch

edfa

ther

PB

BuC

yP

B60

1010

AC

L28

MC

yTB

I11

5M

7H

LA

-mat

ched

sibl

ing

PB

BuC

y60

42N

ML

7P

BC

yTB

I42

47H

LA

-mat

ched

UN

RB

MB

uCyA

TG

BM

60A

LL

512

M24

PB

FB

EA

C15

25H

LA

-mat

ched

sibl

ing

PB

Flu

dara

ML

F90

20H

D23

1327

PB

BE

AC

217

HL

A-m

atch

edsi

blin

gP

BF

luda

raM

LF

80M

1432

HD

aK

arno

fsky

PS,

Kar

nofs

kype

rfor

man

cest

atus

;A

ML

,ac

ute

mie

lobl

asti

cle

ukem

ia;

CM

L,

chro

nic

mie

loid

.le

ukem

ia;

NH

L,

non-

Hod

gkin

lym

phom

a;M

M,

mul

tipl

em

yelo

ma;

HD

,H

odgk

indi

seas

e;M

CL

,man

tle

cell

lym

phom

a;A

LL

,acu

tely

mph

obla

stic

leuk

emia

;Cy,

cycl

opho

spha

mid

e;T

BI,

tota

lbod

yir

radi

atio

n;B

u,bu

sulf

an;B

EA

C,B

CN

U,e

topo

side

,ara

-C,c

yclo

phos

pham

ide;

ML

F,

mel

phal

an;

VP

,et

opos

ide,

Flu

dara

,flu

dara

bine

;U

NR

,un

rela

ted;

BM

,bo

nem

arro

w;

PB

,pe

riph

eral

bloo

d.b

Mon

ths.

C. Martınez et al. / Leukemia Research 25 (2001) 379–384382

Table 2Allogenic transplantation outcome

Chronic GVHD Cause of deathPatient StatusAcute GVHD

01 Grade III NEa Idiopathic pneumonitis+acute GVHD Dead, day +45Extensive Bacterial sepsis+chronic GVHDGrade II Dead, 11 months02

Grade III03 Extensive Chronic GVHD Dead, 11 monthsNE04 Idiopathic pneumonitisGrade I Dead, 2 monthsNE Multiorgan failure+fungal infection0 Dead, day +2105Extensive Chronic GVHD06 Dead, 7 monthsGrade IIExtensive Bacterial sepsis+chronic GVHDGrade II Dead, 9 months07

Grade II08 Extensive Alive, 20 monthsExtensive0 Alive, 20 months09

Grade II10 NV Viral infection Dead, day +107NE Relapse Dead, 3 months11 0NE Viral infection+acute GVHD Dead, 2 monthsGrade II12

Grade III NE13 Alive, 4 months14 NE NE Idiopathic pneumonitis Dead, day +22

a NE, not evaluable.

3.3. Sur6i6al and cause of death

The Kaplan–Meier estimated overall survival at 1year was 16% (median 3.5 months, 95% CI: 0.7–10.3)(Fig. 1). One patient died 3 months after transplant dueto leukemic relapse. Ten patients (71%) died fromtransplant related complications at a median of 3.5months (range 0.7–11) after the procedure. As shownin Table 2, causes of death included GVHD in sixcases; bacterial, fungal or viral infections in five; idio-pathic pneumonitis in three; and multiorgan failure inone. At the time of this report, three patients remainalive (Table 2). Two of them (patients c8 and c9)had mantle cell lymphoma and underwent alloSCT inCR and partial remission, respectively. Both developedextensive chronic GVHD remaining in CR 20 monthsafter transplantation with a Karnofsky performancestatus score of 90%. Patient c13 had a resistantHodgkin disease with lung involvement. She developedan acute grade III GVHD responding to immunosup-pressive therapy. At the time of writing, the patient is inCR, 8 months after the alloSCT and she has developedextensive chronic GVHD. In these patients, the dura-tion of the remission after the second transplant waslonger than after the autoSCT.

4. Discussion

Patients with hematological malignancies who relapseafter autoSCT have a very poor prognosis with amedian survival of less than 1 year [4]. In most cases, asecond autoSCT is not feasible due to the resistance ofthe disease to the therapy, or to the impossibility ofcollecting enough stem cells. For all these reasons,alloSCT is a potentially interesting salvage therapyoffering the advantages of both the lack of tumor

contamination of the graft and the antitumoral activitymediated by the so-called graft-versus-tumor effect[13–16].

The outcome of patients receiving an alloSCT afterfailing an autoSCT has been analyzed in several studies,the majority coming from registries or multiple institu-tions [7–10]. The largest series analyzed treatment re-sults in patients with acute leukemia and has recentlybeen reported by the European Blood and MarrowTransplantation Registry [7]. In this study, patientsrelapsing after an autoSCT underwent a second au-tologous (n=74) or allogenic (n=94) transplant assalvage therapy. Transplanted-related mortality (TRM)at 2 years was 51% in recipients of matched alloSCTand 26% following a second autoSCT (PB0.05).Leukemia-free survival (LFS) was not statistically dif-ferent between the two groups since higher diseaseprogression was observed after autoSCT. In the multi-variate analysis, parameters associated with shorterLFS were: age \25 years, interval from first autoSCT

Fig. 1. Overall survival after alloSCT.

C. Martınez et al. / Leukemia Research 25 (2001) 379–384 383

to relapse of 8 months or less, and TBI used as partof the preparative regimen in the first autoSCT.

The role of second transplants in the managementof patients with lymphoid malignancies has not beenwidely investigated [9,10,17–19]. Tsai et al. [9] ana-lyzed 20 patients, 14 of them with lymphoma, whounderwent alloSCT after failing autoSCT. In thisstudy, the median survival was only 2 months; all thedeaths were due to transplant-related complicationsthat included bacterial or fungal infection, veno-occlu-sive disease of the liver, idiopathic interstitial pneu-monitis and chronic GVHD. Only three patients werealive at the time of this report. Similar results havebeen shown in patients with multiple myeloma havingrelapsed after autoSCT [10,20]. Thus, Mehta et al.[10] performed a case-matched study comparing pa-tients who underwent an alloSCT or a second au-toSCT for relapsed or refractory multiple myeloma.The conclusions of this study were that while au-toSCT might be superior to alloSCT when consider-ing overall survival (54 vs. 29%, respectively,P=0.01), the event-free survival was comparablegiven the significantly lower disease progression afteralloSCT (72 vs. 31%, respectively, P=0.03).

In our single institution study, which included allpatients that had received an alloSCT after havingrelapsed from an autoSCT, alloSCT resulted inprompt engraftment. However, TRM due to toxicity,and the incidence of GVHD and infections were ex-tremely high. Although the series is small, the propor-tion of long-term survivors does not appear to bedifferent from that found in other series. It should benoted that, although most patients had relapsed after8 months from autoSCT and only four received TBIas part of the preparation regimen for the first trans-plant, a great proportion of them were over 25 andhad been heavily pretreated, factors which have beenassociated with poor outcome [7–10]. The last twopatients of the series received low-intensity condition-ing regimens in order to reduce the toxicity associatedto the procedure. One of them is alive and in CR atthe time of this report

In summary, in patients who relapse after an au-toSCT, salvage therapy with alloSCT is associatedwith a high TRM, but a proportion of patients mayexperience prolonged disease-free survival. The mostimportant problem with this procedure is TRM. Thiscould be diminished by a careful selection of the pa-tients and, particularly, by reducing the intensity ofthe conditioning regimen, an avenue that is being ac-tively investigated [21,22]. Taking into account thedismal prognosis of patients relapsing after a firsttransplant, as well as the high TRM associated withsecond regular transplants, the so-called ‘mini-trans-plants’ (or transplants with non-myeloablative regi-mens) are worth investigation.

Acknowledgements

This work has been supported in part by grantsSGR 1998/00111 from the Comissionat per a Univer-sitats i Recerca (Generalitat Catalunya), FISS-98/995,98/0380, and 99/0189 from the Fondo de Investiga-ciones Sanitarias de la Seguridad Social, SpanishMinistry of Health, and FIJC-00/P-CR and FIJC-00/P-EM from the Jose Carreras International Founda-tion. We wish to thank Anna Bauer for the Englishgrammar review of the manuscript.

References

[1] Foon KA, Gale RP. Therapy of acute myelogenous leukemia.Blood Rev 1992;6:15.

[2] Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bonemarrow transplantation as compared with salvage chemotherapyin relapses of chemotherapy-sensitive non-Hodgkin’s lymphoma.New Engl J Med 1995;333:1540.

[3] Haioun C, Lepage E, Gisselbrecht C, et al. Comparison ofautologous bone marrow transplantation with sequentialchemotherapy for intermediate-grade and high-grade non-Hodgkin’s lymphoma in first complete remission: a study of 464patients. J Clin Oncol 1994;12:2543.

[4] Vose JM, Bierman PJ, Anderson JR, et al. Progressive diseaseafter high-dose chemotherapy and autologous marrow transplan-tation for lymphoid malignancy: clinical course and patientfollow-up. Blood 1992;80:2142.

[5] Mrsıc M, Horowitz MM, Atkinson K, et al. Second HLA-identi-cal sibling transplants for leukemia recurrence. Bone MarrowTransplant 1992;9:269.

[6] Barret AJ, Locatelli F, Treleaven JG, Gratwohl A, Szydlo R,Zwaan FE. Second transplants for leukaemic relapse after bonemarrow transplantation: high early mortality but favourableeffect of chronic GVHD on continued remission. Br J Haematol1991;79:567.

[7] Ringden O, Labopin M, Frassoni F, et al. Allogeneic bonemarrow transplant or second autograft in patients with acuteleukemia who relapse after an autograft. Bone Marrow Trans-plant 1999;24:389.

[8] Blau IW, Basara N, Bischoff M, et al. Second allogeneic hemato-poietic stem cell transplantation as treatment for leukemia re-lapsing following a first transplant. Bone Marrow Transplant2000;25:41.

[9] Tsai T, Goodman S, Saez R, et al. Allogeneic bone marrowtransplantation in patients who relapse after autologous trans-plantation. Bone Marrow Transplant 1997;20:859.

[10] Mehta J, Tricot G, Jagannath S, et al. Salvage autologous orallogeneic transplantation for multiple myeloma refractory to orrelapsing after a first-line autograft? Bone Marrow Transplant1998;21:887.

[11] Martınez C, Carreras E, Rovira M, et al. Patients with mantle-cell lymphoma relapsing after autologous stem cell transplanta-tion may be rescued by allogeneic transplantation. Bone MarrowTransplant 2000;26:677–9.

[12] Thomas ED, Storb R, Clift RA, et al. Bone marrow transplanta-tion. New Engl J Med 1975;292:832.

[13] Horowitz MM, Gale RP, Sondel PM, et al. Graft-versus-leukemia reactions after bone marrow transplantation. Blood1990;75:555.

[14] Jones RJ, Ambinder RF, Piantadosi S, Santos DW. Evidence ofa graft-versus-lymphoma effect associated with allogeneic bonemarrow transplantation. Blood 1991;77:649.

C. Martınez et al. / Leukemia Research 25 (2001) 379–384384

[15] Verdonck LF, Dekker AW, Lokhorst HM, Petersen EJ,Nieuwenhuis HK. Allogeneic versus autologous bone marrowtransplantation for refractory and recurrent low-grade non-Hodgkin’s lymphoma. Blood 1997;90:4201.

[16] van Besien KW, de Lima M, Giralt SA, et al. Management oflymphoma recurrence after allogeneic transplantation: the rele-vance of graft-versus-lymphoma effect. Bone Marrow Transplant1997;19:977.

[17] Schouten HC, Armitage JO, Klassen LW, et al. Allogeneic bonemarrow transplantation in patients with lymphoma relapsingafter autologous marrow transplantation. Bone Marrow Trans-plant 1989;4:119.

[18] Moreau P, Mechinaud F, Mahe B, et al. Successful allogeneicbone marrow transplantation for early relapse after autologousbone marrow transplantation in two cases of aggressive high-grade non-Hodgkin’s lymphoma. Bone Marrow Transplant1996;18:665.

[19] de Lima M, van Besien KW, Giralt SA, et al. Bone marrowtransplantation after failure of autologous transplant for non-Hodgkin’s lymphoma. Bone Marrow Transplant 1997;19:121.

[20] Toor A, Morris C, Spencer H, et al. Allogeneic stem celltransplantation as salvage or consolidation therapy after priormelphalan based autotransplantation for poor risk multiplemyeloma. Blood 1999;10(Suppl. 1):346a.

[21] Slavin S, Nagler A, Naparstek E, et al. Nonmyeloablative stemcell transplantation and cell therapy as an alternative to conven-tional bone marrow transplantation with lethal cytoreduction forthe treatment of malignant and nonmalignant hematologic dis-eases. Blood 1998;91:756.

[22] Khouri IF, Keating M, Korbling M, et al. Transplant-lite:induction of graft-versus-malignancy using fludarabine-basednonablative chemotherapy and allogeneic blood progenitor-celltransplantation as treatment for lymphoid malignancies. J ClinOncol 1998;16:2817.

.