Allogeneic haematopoietic stem cell transplantation for the treatment of adult T-cell leukaemia/lymphoma


Dr Masahiro Kami, Department of Medical Oncology, Haematopoietic Stem Cell Transplantation Unit, National Cancer Centre Hospital, 5-1-1 Tsjukiji, Chuo-ku, Tokyo 104–0045, Japan. E-mail:


Summary. The feasibility of allogeneic haematopoietic stem-cell transplantation (allo-HSCT) in 11 patients with adult T-cell leukaemia/lymphoma (ATL) (6 acute, 4 lymphoma, 1 chronic type) was evaluated. The preparative regimens (9 conventional, 2 reduced-intensity) were tolerable. Five patients developed acute graft-versus-host disease (GVHD), and three, extensive chronic GVHD. All 10 patients who survived > 30 d achieved complete remission. Estimated 1-year overall and disease-free survival rates were 53 ± 30% and 45 ± 29% respectively. Four patients remain alive and disease-free at a median follow-up of 25 months. The others died of transplantation-related complications. This pilot study suggests that allo-HSCT in ATL should be evaluated further.

Adult T-cell leukaemia/lymphoma (ATL) is one of the most aggressive haematological malignancies associated with human T-cell lymphotropic virus type I (HTLV-I). Both acute and lymphoma types of ATL usually have a poor prognosis with survival of 6 to 13 months (Taguchi et al, 1996; Yamada et al, 2001).

Successful allogeneic bone marrow transplantation (BMT) (allogeneic haematopoietic stem-cell transplantation, allo-HSCT) has been reported (Borg et al, 1996; Tajima et al, 2000). Interestingly, the median leukaemia-free survival in 10 ATL patients after BMT was 17·5 months (Utsunomiya et al, 2001), which was much longer than that previously reported, suggesting that allo-HSCT might prove a promising measure.

We herein describe our feasibility study of allo-HSCT to treat ATL and discuss the clinical usefulness of this treatment.

Patients and methods

Diagnosis and classification of clinical subtypes of ATL.  The diagnostic criteria were as follows: (1) anti-HTLV-I positive serum; (2) abnormal lymphocytes with convoluted or lobulated nuclei in the peripheral blood, or histological findings compatible with ATL in biopsied lymph nodes; (3) tumour cells with mature CD4+ T-cell phenotype; and (4) clonal integration of HTLV-I proviruses in the DNA of tumour cells determined by Southern blotting. The clinical subtypes have been reported (Shimoyama, 1991).

Patients' characteristics.  In total, 11 patients (acute (n = 5), lymphoma (n = 4) and chronic type (n = 2)) underwent allo-HSCT from related donors (nine human leucocyte antigen (HLA)-identical and one two-loci mismatched) or a matched unrelated donor (n = 1) between April 1999 and April 2001 (Table I); five were born in Kyusyu, an endemic area of HTLV-1 infection in Japan (Tajima, 1988). Disease status at transplantation was complete remission in six, partial remission in one and no response in four. The number of pretransplant chemotherapeutic regimens was two (range 1–3).

Table I.  Characteristics and outcomes of 11 patients with adult T-cell leukaemia undergoing allogeneic haematopoietic stem cell transplantation (HSCT). : A. Cases 1–6.
Patients' status
Age (years)/sex47/M46/F59/M48/F27/F46/F
Type of ATLAcuteAcuteLymphomaChronicAcuteAcute
Timing of
1st complete remission1st complete remissionProgressive disease1st relapse1st complete remissionProgressive disease
Liver dysfunction due to
hepatitis C virus
NoneLiver dysfunction due
to hepatitis B virus
NoneNoneHepatitis C
virus infection
Donors' status
Source of stem
Bone marrowBone marrowBone marrowG-CSF-mobilized
peripheral blood
Bone marrowG-CSF-mobilized
peripheral blood
DonorHLA-identical siblingHLA-identical siblingMatched unrelated
HLA-identical siblingHLA-identical siblingHLA-identical
HTLV-I antibodyNegativePositiveNegativeNegativeNegativePositive
Transplantation proceedings
CY, TBICytarabine, CY, TBIFludarabine, CY, ATGCY, TBICY, TBIBU, CY
sMTX and cyclosporinesMTX and cyclosporineCyclosporine alonesMTX and cyclosporinesMTX and cyclosporinesMTX and cyclosporine
cells (108/kg)
0·823·42·5 3·84·1
CD34+cells (106/kg)
> 0·5 × 109/l (day)
Platelets >
20 × 109/l (day)
> 1% (day)
(> grade II)
(each grade II)
Liver, kidneyNoneNone toxicityLung, kidney
(each grade III)
VODNoneEsophageal bleedingNoneNoneHaemorrhagic
cystitis due
to BK virus
InfectionCMV enterocolitisNoneCMV antigenaemiaCMV antigenaemiaNoneInvasive aspergillosis
CMV gastritis,
(S. maltophilia)
Chronic GVHDNoneNoneNot evaluableExtensiveNoneExtensive
Overall survival
From onset
13·026·8 +2·315·031·5 +4·5
Event-free survival
13·026·8+2·32·031·5 +4·5
Causes of deathChronic GVHDAliveGastrointestinal bleedingChronic GVHDAliveGastrointestinal
bleeding due to
acute GVHD

Preparative regimen and graft-versus-host disease (GVHD) prophylaxis.  In total, nine patients received conventional BMT (four cyclophosphamide/total body irradiation (CY/TBI), two cytarabine (Ara-C)/CY/TBI, two etoposide (VP-16)/CY/TBI, one busulphan (BU)/CY) and GVHD prophylaxis of cyclosporine and short-term methotrexate. Two received a reduced-intensity regimen; fludarabine/CY with or without anti-thymocyte globulin (ATG), and cyclosporine alone. The stem cells were not manipulated.

Response to treatment.  The response criteria were defined as follows (Gill et al, 1995): complete response (resolution of all malignant disease for 4 weeks or more); partial remission (reduction in measurable indices by lasting 4 weeks or more, without the development of new lesions or disease progression); and progressive disease (increase in measurable disease or in the number of circulating leukaemic cells of 25% or more).

Statistical analysis.  The actuarial probability of survival and the cumulative probability of a response were calculated with the Kaplan–Meier method. Univariate analysis using Fisher's exact test was performed to evaluate the risk of non-relapse mortality.



All patients tolerated the preparative regimen and achieved neutrophil recovery (> 0·5 × 109/l); however, five patients died before platelet recovery. In the remaining six patients, it took 26·5 d (range 23–38) for the platelet count to exceed 50 × 109/µl.

GVHD and response to allo-HSCT

Five patients developed acute GVHD and three out of the eight patients who survived > 100 d developed extensive chronic GVHD (Table I). Complete remission was observed in the 10 patients who survived > 30 d. Two patients experienced leukaemic recurrence, and achieved complete remission with the rapid withdrawal of cyclosporine and donor lymphocyte infusion (cases 4 and 10). They died of chronic GVHD 15·0 and 7·7 months after transplantation.


Five patients developed grade II to III regimen-related toxicity (Bearman et al, 1988) (Table I). Infectious complications were observed in seven patients. Non-infectious complications included hepatic veno-occlusive disease, interstitial pneumonitis and esophageal bleeding (n = 1, respectively).


Seven patients died of transplantation-related complications including acute GVHD (n = 2), chronic GVHD (n = 3), gastrointestinal bleeding (n = 1) and interstitial pneumonitis (n = 1). The estimated 1-year overall and disease-free survival rates were 54·5% ± 30·0% and 45 ± 30·0% respectively. Currently, four patients are alive without either acute or chronic GVHD and disease-free at a median follow-up of 25 months (range, 12–32).

Four out of the six patients who were transplanted in complete remission survived, whereas all of the other five patients died of GVHD or regimen-related toxicity (P = 0·0606).


Adult T-cell leukaemia/lymphoma is a highly aggressive haematological malignancy. With the exception of some small studies (Gill et al, 1995; Yamada et al, 2001), its prognosis is poor with median survival rates ranging from 5·5 to 8·5 months (Pawson et al, 1998; Matsushita et al, 1999). Without an established curative treatment (Tsukasaki et al, 1999), our results are encouraging. Four patients have survived for a median of 25 months, and the 1-year overall survival was 54·5%. Furthermore, all patients achieved complete remission, and those who relapsed responded to donor lymphocyte infusion. These findings suggest that a graft-versus-ATL effect does exist and allo-HSCT is promising.

However, transplantation-related mortality was a significant problem in this study. Seven patients died without ATL progression of acute GVHD (n = 2), chronic GVHD (n = 3) and regimen-related toxicity (n = 2).

A graft-versus-leukaemia (GVL) effect is essential for eradicating ATL cells. Importantly, however, the survivors did not develop acute or chronic GVHD. GVHD might not be essential for a GVL effect or disease-free survival in ATL. GVHD prophylaxis warrants further investigation.

Two patients died of regimen-related toxicity, and five patients developed documented infections. Long-term survival was limited to the patients in first remission. As ATL was well controlled in all the patients after HSCT, accumulated organ damage as a result of repeated cytotoxic chemotherapies might have contributed to their deaths. Heavily treated patients are not good candidates for HSCT. Reducing these complications by the reduced-intensity transplant approach will be the focus of future investigation.

The study size was too small to reach a conclusion, and the exact mechanism of the GVL effect in ATL remains unknown. However, our data indicates that allo-HSCT may offer the best chance of haematological remission and prolonged survival for patients with ATL with a good performance status. Allo-HSCT should continue to be examined in the context of clinical trials and be directed at developing approaches to minimize toxicity and prevent GVHD.