Allogeneic hematopoetic stem-cell transplantation for patients with relapsed or refractory lymphomas: comparison of high-dose conventional conditioning versus fludarabine-based reduced-intensity regimens

H. Bertz, G. Illerhaus, , H. Veelken and J. Finke,+

Albert Ludwigs University Medical Center Freiburg, Department of Haematology and Oncology, Freiburg, Germany

Received 23 November 2000; revised 2 July 2001; accepted 14 August 2001.


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
Background

Allogeneic hematopoetic stem-cell transplantation (alloHSCT) has curative potential for poor risk lymphoma patients due to the graft-versus-lymphoma effect. High non-relapse mortality with conventional high-dose conditioning indicates the necessity for less toxic transplant strategies.

Patients and methods

Between 1992 and 1999, 25 patients [median age 37 (20–60) years] with relapsed or refractory non-Hodgkin’s lymphoma (NHL, n = 20) or Hodgkin’s disease (HD, n = 5) received an alloHSCT in our institution. Patients were grafted from HLA matched (17) or mismatched (2) related, or matched unrelated donors (MUD) (6). NHL histological subtypes were lymphoblastic (6), high grade B/T-cell lymphomas (5), follicular (3), mantle cell (2) and CLL, immunocytic, composite lymphoma and panniculitic T-NHL in one patient each. Patients had received a median of four (range three to six) different therapies before alloHSCT, and 10 patients had relapsed after high-dose chemotherapy and autologous (9) or allogeneic (1) HSCT. Remission status prior to allogeneic SCT was CR1 (1), CR2 (1), relapse (11), partial remission (5) or primary refractory induction failure (7). Conventional myeloablative conditioning (cc) regimens contained total body irradiation 12 Gy (5), busulfan 16 mg/kg (7) or BCNU/VP16 (1). Twelve patients received reduced-intensity conditioning (ric) regimens with fludarabine (FLU) plus alkylating agents. Graft-versus-host disease prophylaxis consisted of cyclosporin A ± prednisone or methotrexate. Six patients also received anti-T-lymphocyte globulin.

Results

Twenty-four patients engrafted. Best response after alloHSCT was complete remission in 16 of all patients [64%; 95% confidence interval (CI) 44% to 84%] and in 16 of 22 evaluable patients (73%; 95% CI 53% to 93%), partial remission in three of 25 (12%), and no change in three of 25 (12%) patients. Early death prevented response evaluation in three of 25 patients. Non-relapse mortality was 54% (95% CI 15% to 78%) in patients after cc and 17% (95% CI 0% to 41%) after FLU-based ric (P = 0.03). Six patients died due to progressive disease or relapse. Four patients with HD died, three in complete remission due to non-relapse mortality and one with progressive disease. Eleven of 25 patients are alive with a median follow up of 618 days (range 383–2815), with an overall survival of 44% (95% CI 23% to 65%) at 1 year for all patients, while eight of 12 (67%; 95% CI 35% to 98%) patients are alive after ric compared with three of 13 (23%; 95% CI 0% to 50%) after cc (P <0.02).

Conclusions

AlloHSCT induces high rates of complete remission in advanced lymphoma patients, even when the tumor had relapsed after autologous HSCT. It should be considered earlier as part of the therapeutic options in poor risk patients to avoid non-relapse mortality associated with extensive pretreatment. Our novel reduced conditioning regimens show promising results, especially in heavily pretreated patients, and improve survival after allogeneic transplantation.

Key words: allogeneic HSCT, Hodgkin’s disease, MUD transplantation, non-Hodgkin’s lymphoma


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
The curative potential of allogeneic hematopoetic stem-cell transplantation (alloHSCT) for advanced lymphoid malignancies has been demonstrated [1, 2], including low-grade (e.g. CLL) and mantle cell lymphoma [3], refractory or relapsed follicular and indolent non-Hodgkin’s lymphomas (NHLs) [1] or Hodgkin’s disease (HD) [4]. The graft-versus-lymphoma effect appears to be particularly important for a reduced incidence of relapse after alloHSCT [2, 5, 6]. The advantage in overall survival achieved by a lower relapse rate after alloHSCT compared with autologous HSCT (autoHSCT) is jeopardized by a high treatment-related morbidity and mortality (non-relapse mortality). The main causes are organ toxicity, graft-versus-host disease (GvHD) and infection. The incidence of these problems is increased in heavily pretreated patients [7]. In an attempt to reduce non-relapse mortality, we have developed reduced-intensity conditioning (ric) regimens [8]. To estimate the potential advantage of these regimens, we retrospectively analyzed the outcome for 25 consecutive patients with lymphomas receiving alloHSCT in our institution between 1992 and 1999. Sixteen of all patients [64%; 95% confidence interval (CI) 44% to 84%] and 16 of 22 evaluable patients (73%; 95% CI 53% to 93%) achieved complete remission. To date (1 November, 2000), 11 of 25 patients (44%; 95% CI 23% to 65%) are alive, with a median follow up of 618 days (range 383–2815). Furthermore, we compared different regimens in chemoresistent and chemosensitive lymphomas.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
Patients, donors and grafts
The patient characteristics, histological diagnoses and remission status at transplantation are shown in Table 1. All 25 consecutive lymphoma patients treated by alloHSCT during the period from January 1992 to October 1999 were included in the analysis. In some cases patients had been heavily pretreated with conventional chemotherapy and radiotherapy: a median of four (range three to six) previous regimens. Nine patients (36%) had relapsed after autoHSCT and one after alloHSCT. The patients transplanted in first complete remission (CR1) had high-risk Burkitt’s lymphoma and had failed to achieve complete remission after standard induction therapy. The patient with lymphoblastic lymphoma, who was transplanted in CR2, had previously relapsed at an extranodal site. All HD patients had previously received radiotherapy, and two had relapsed after autoHSCT. Grafts were from a matched or mismatched sibling/relative or twin in 19 patients, and from a matched unrelated donor (MUD) in six transplantations.


View this table:
[in this window]
[in a new window]
 
Table 1. Patients, donors and graft characteristics
 
Conditioning regimens and GvHD prophylaxis
Conventional conditioning (cc) regimens consisted of total body irradiation (TBI) 12 Gy or busulfan 16 mg/kg + cyclophosphamide ± VP16 in nine of 25 patients (36%); individually designed protocols were used in 16% of patients (Table 1). To reduce non-relapse mortality, in 1998 we introduced a fludarabine (FLU)-based ric regimen (FBM) [8], containing 150 mg/m2 FLU, 400 mg/m2 carmustine (BCNU) and 140 mg/m2 melphalan (MEL). In patients >55 years of age, BCNU and MEL were reduced to 300 and 110 mg/m2, respectively. Nine patients received the FBM regimen. Three patients, who had previously failed autoHSCT, were treated with a modified protocol for patients with contraindications or extensive pretreatment with alkylating agents. They received thiotepa (TT) 3 x 5 mg/kg, replacing BCNU and MEL. Patients receiving FBM were treated within an open-label phase II trial approved by the local internal review board (IRB) and ethics committee; informed consent was given. The trial was carried out according to the Declaration of Helsinki.

GvHD prophylaxis consisted of cyclosporin A ± methotrexate or prednisone. Anti-T-lymphocyte globulin (ATG-S®; Fesenius Graefeling, Germany) (n = 6) was added in patients with a MUD (n = 5) or mismatched related donor (n = 1).

Supportive care
Supportive care was carried out as published previously [9]. All patients received G-CSF starting on day 1 (n = 5) or day 7 (n = 20).

Statistical analysis
The Kaplan–Meier method was used to calculate event-free survival, overall survival, the probability of non-relapse mortality and the CI, using GraphPad Prism software (San Diego, CA, USA).


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
Engraftment
One patient died on day 20 before engraftment. In 24 of 25 patients, the median day of definitive engraftment with leukocytes >=1 x 109/l was day 12 (range 10–18); platelet recovery >=20/>=50 x 109/l was reached in 21 and 19 patients in median day 14 (range 9–43) and day 18 (range 12–104), respectively. Four patients died before stable platelet engraftment occurred.

Graft-versus-host disease
Acute GvHD °0/I developed in 76% (95% CI 58% to 94%), °II–IV in 24% (95% CI 6% to 42%), and °III–IV in 16% (95% CI 1% to 31%) of all patients. In 16 patients surviving beyond day 100, chronic GvHD is absent (n = 8), limited (n = 5) or extensive (n = 3).

Outcome
Response is evaluable in 22 of 25 patients. Sixteen patients (73%) achieved complete remission and complete donor chimerism in the bone marrow. One patient with composite lymphoma, after previous autoHSCT, relapsed on day 43. A secondary sustained complete remission was achieved after cessation of immunosuppression in conjunction with local radiation therapy. As of 1 November, 2000, 11 of 25 are alive at a median follow-up of 618 days (range 383–2815), with an overall survival at 1 year of 44% (95% CI 23% to 65%) (Figure 1). Four patients achieving only short-term partial remission died of progressive disease on days 31, 35, 82 and 162, respectively. None of these four patients had ever achieved a remission after any standard chemotherapy. Two patients, also after autoHSCT, relapsed on days 75 and 123, and died due to progressive disease on days 159 and 267, respectively. ARDS with respiratory failure was the cause of death in three patients with previous TBI (12 Gy) or mediastinal bulk irradiation. Other causes of death were interstitial pneumonia (two patients), aspergillosis (2) and hemolytic uremic syndrome/microangiopathic hemolytic anemia (HUS/MAHA) (1). Four of five patients with HD died because of ARDS (two patients), HUS/MAHA (1) in complete recovery, or persistent disease (1). Overall non-relapse mortality in all patients was 32% (95% CI 12% to 52%). Event-free survival, is 36% (95% CI 16% to 50%) for all patients at 1 year.



View larger version (20K):
[in this window]
[in a new window]
 
Figure 1. Overall survival (OS) in alloHSCT for poor risk lymphoma patients (n = 25) and comparison of risk of non-relapse mortality (TRM) for patients with FLU-based ric (n = 12) and TBI/busulfan- or BCNU/VP16-containing cc regimens (n = 13) (P = 0.03).

 
Ric versus cc
In the subgroup of patients receiving FLU/BCNU/MEL or FLU/TT, 67% (95% CI 35–98%) are alive with a median survival of 575 days (range 383–805) and seven of 12 (60%) are in complete remission. In contrast, overall survival and event-free survival in the group treated with cc is only 23% (95% CI 0% to 50%) and 23% (95% CI 0% to 50%), respectively. In the FLU subgroup, two patients died of relapse and two patients in complete remission died due to viral pneumonitis during intensive immunosuppression. Despite the overall small numbers of patients, statistical analysis demonstrated superior overall survival and event-free survival in the ric over the cc regimens (67% versus 23%, P <0.02; and 50% versus 23%, P = 0.04, respectively). In addition, non-relapse mortality with ric compared with cc (17% versus 54%) is statistically significant (P = 0.03) (Figure 1).

Chemoresistent versus chemosensitive lymphomas
Twenty-three patients, not transplanted in complete remission, were divided into the subgroups chemoresistent (13) and chemosensitive (10) lymphomas. Three of 13 in the first group (23%; 95% CI 0% to 50%) are alive and six of 13 (46%) achieved complete recovery compared with six of 10 (60%; 95% CI 23% to 97%) and nine of 10 (90%) in the chemosensitive group. Non-relapse mortality is 31% (95% CI 2% to 60%) and 40% (95% CI 6% to 77%), and event-free survival 23% (95% CI 0% to 50%) versus 40% (95% CI 6% to 77%); differences are not statistically significant. Overall survival shows an advantage for the chemosensitive group (23% versus 60%; not statistically significant).

Patients with and without previous autoHSCT
Seven of nine patients with previous autoHSCT achieved complete remission (78%) and four of nine (44%; 95% CI 4% to 85%) are alive compared with seven of 15 (47%; 95% CI 18% to 60%) without previous transplantation. Overall survival (44% versus 47%), event-free survival (22% versus 47%) and non-relapse mortality (33% versus 27%) showed no statistically significant differences.


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
Long-term survival in patients with relapsed or refractory NHL/HD remains poor, even with salvage chemo- or radiotherapy [10]. The graft-versus-lymphoma effect shown in lymphomas [1, 2, 11] and HD [6] might make alloHSCT a potential curative approach. Several studies have described a better overall survival and disease-free survival in younger patients [2], or in patients with low-grade NHL [1] treated with upfront alloHSCT. We report the results of alloHSCT in the treatment of 25 poor risk patients, who had already received extensive pretreatment, including nine patients (36%) with relapse after autoHSCT. Even after autoHSCT, alloHSCT can be a curative salvage therapy, as shown by de Lima et al. [5]. The main problem in heavily pretreated patients is a high non-relapse mortality of 30% to 65% [1, 5, 6], mainly associated with infections, GvHD [5] and idiopathic or interstitial pneumonia due to previous radiotherapy [7]. To reduce non-relapse mortality, we have developed a FLU-based ric regimen [8]. With this approach, we observed a non-relapse mortality of 17% (95% CI 0% to 41%) (two of 12 patients dying on days 107 and 243). This low non-relapse mortality is in accordance with recently published data [12] and compares favorably with the non-relapse mortality of 54% (95% CI 15% to 78%) (six of 13) with cc regimens (P = 0.03). Conversely, six of eight patients, who died of aspergillosis, interstitial pneumonia, MAHA or ARDS, had received high-dose cc.

The very poor risk of our patient population must be taken into account when the efficacy of alloHSCT for NHL is being considered. Only 8% were transplanted in complete recovery, but 64% of all patients and 73% of the evaluable patients achieved complete remission with alloHSCT. Twelve patients have not relapsed and seven patients with hitherto chemoresistant relapsed lymphoma or primary induction failure achieved complete recovery. Seven of eight evaluable patients who had received previous autoHSCT achieved complete remission (87%) and only two of seven have relapsed. Therefore, failure of previous autoHSCT does not appear to adversely affect the treatment results or non-relapse mortality after alloHSCT.

Our results are certainly encouraging, but the outcome of the patients with HD after cc is disappointing. Three patients died in complete remission due to non-relapse mortality, mainly ARDS, and one failed to achieve a response. Only one patient is alive in partial remission as of day 383. This poor outcome has been observed previously by Milpied et al. [6], who reported a non-relapse mortality of 65%. The most likely explanation may be found in the extensive pretreatment including radiotherapy. We agree with Anderson [4] that only selected patients with HD should receive alloHSCT and reduced intensity conditioning regimens should be introduced earlier in high-risk patients.

In conclusion, our results appear to support the assumption that alloHSCT has curative potential in patients with poor risk lymphomas, most likely due to the lack of lymphoma involvement of the graft and the proven graft-versus-lymphoma effect. Significantly reduced non-relapse mortality can be expected from the use of FLU-based ric regimens without loss of efficacy. Reduced non-relapse mortality may also permit incorporation of alloHSCT earlier into the therapy course in patients with unfavorable entities such as mantle cell lymphomas [13].


    Acknowledgements
 
We thank I. Matt for the excellent documentation work, E. Lenartz for coordination, L. Veelken-Ahn for language assistance, and S. Wider, E. Zamek, S. Zehbe and U. Tritschler for technical assistance.


    Note added in proof
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
As of 1 November 2001, 11 of the 25 patients are alive with a median follow-up of 1021 days (range 748–3180). No further relapse has ocurred.


    Footnotes
 
+ Correspondence to: Hugstetter Str. 55, 79106 Freiburg, Germany. Tel: +49-761-270-3321; Fax: +49-761-270-3233; E-mail: Finke@mm11.ukl.uni-freiburg.de Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 Note added in proof
 References
 
1. Verdonck LF. Allogeneic versus autologous bone marrow transplantation for refractory and recurrent low-grade non-Hodgkin’s lymphoma: updated results of the Utrecht experience. Leuk Lymph 1999; 34: 129–136.[ISI][Medline]

2. Ratanatharathorn V, Uberti J, Karanes C et al. Prospective comparative trial of autologous versus allogeneic bone marrow transplantation in patients with non-Hodgkin’s lymphoma. Blood 1994; 84: 1050–1055.[Abstract/Free Full Text]

3. van Besien K, Sobocinski KA, Rowlings PA et al. Allogeneic bone marrow transplantation for low-grade lymphoma. Blood 1998; 92: 1832–1836.[Abstract/Free Full Text]

4. Anderson JE, Litzow MR, Appelbaum FR et al. Allogeneic, syngeneic, and autologous marrow transplantation for Hodgkin’s disease: the 21-year Seattle experience. J Clin Oncol 1993; 11: 2342–2350.[Abstract]

5. de Lima M, van Besien KW, Giralt SA et al. Bone marrow transplantation after failure of autologous transplant for non-Hodgkin’s lymphoma. Bone Marrow Transplant 1997; 19: 121–127.[ISI][Medline]

6. Milpied N, Fielding AK, Pearce RM et al. Allogeneic bone marrow transplant is not better than autologous transplant for patients with relapsed Hodgkin’s disease. European Group for Blood and Bone Marrow Transplantation. J Clin Oncol 1996; 14: 1291–1296.

7. Bearman SI, Appelbaum FR, Back A et al. Regimen-related toxicity and early posttransplant survival in patients undergoing marrow transplantation for lymphoma. J Clin Oncol 1989; 7: 1288–1294.[Abstract]

8. Wäsch R, Reisser S, Hahn J et al. Rapid achievement of complete donor chimerism and low regimen-related toxicity after reduced conditioning with fludarabine, carmustine, melphalane and allogeneic transplantation. Bone Marrow Transplant 2000; 26: 243–250.[ISI][Medline]

9. Bertz H, Potthoff K, Mertelsmann R, Finke J. Busulfan/cyclophosphamide in volunteer unrelated donor (VUD) BMT: excellent feasibility and low incidence of treatment related toxicity. Bone Marrow Transplant 1997; 19: 1169–1173.[ISI][Medline]

10. Philip T, Armitage JO, Spitzer G et al. High-dose therapy and autologous bone marrow transplantation after failure of conventional chemotherapy in adults with intermediate-grade or high-grade non-Hodgkin’s lymphoma. N Engl J Med 1987; 316: 1493–1498.[Abstract]

11. van Besien KW, de Lima M, Giralt SA et al. Management of lymphoma recurrence after allogeneic transplantation: the relevance of graft-versus-lymphoma effect. Bone Marrow Transplant 1997; 19: 977–982.[ISI][Medline]

12. Nagler A, Slavin S, Varadi G et al. Allogeneic peripheral blood stem-cell transplantation using a fludarabine-based low intensity conditioning regimen for malignant lymphoma. Bone Marrow Transplant 2000; 25: 1021–1028.[ISI][Medline]

13. Khouri IF, Lee M, Romaguera J et al. Allogeneic hematopoetic transplantation for mantle-cell lymphoma: Molecular remissions and evidence of graft-versus-malignancy. Ann Oncol 1999; 10: 1293–1299.[Abstract]