Effect of Age and Previous Autologous Transplantation on Nonrelapse Mortality and Survival in Patients Treated With Reduced-Intensity Condit
http://www.100md.com
《临床肿瘤学》
the Divisione di Ematologia and Ufficio Operativo, Istituto Nazionale Tumori, Università di Milano
Divisione di Ematologia, Unità Trapianto Midollo, Ospedale San Raffaele, Milano
Divisione di Ematologia II, Ospedale San Martino, Genova
Divisione di Ematologia, Università di Ancona, Ancona
Divisione di Ematologia, Università di Udine, Udine
Divisione di Ematologia, Ospedale Riuniti, Bergamo
Unità Trapianto Midollo, Ospedale Cervello, Palermo
Divisione di Ematologia, Ospedale San Camillo-Forlanini, Roma
Istituto di Ematologia–Oncologia, Università di Bologna, Bologna
Divisione di Ematologia, Università di Torino, Torino, Italy
ABSTRACT
PURPOSE: Older age and a previously failed autologous stem-cell transplantation (SCT) are poor prognostic factors for patients receiving myeloablative conditioning and allogeneic SCT. Reduced-intensity conditioning (RIC) regimens achieved a significant reduction of treatment-related mortality, but the influence of previously described risk factors on the outcome of this novel transplantation strategy have not been fully analyzed yet.
PATIENTS AND METHODS: One hundred fifty patients with advanced hematologic malignancies received a RIC regimen containing thiotepa (10 mg/kg), fludarabine (60 mg/m2), and cyclophosphamide (60 mg/kg), followed by an allogeneic transplantation from an HLA-identical sibling donor. Patients were divided into two cohorts according to age; 90 patients were younger than 55 years, and 60 patients were 55 years old or older. The other pretransplantation characteristics were fairly balanced.
RESULTS: Actuarial 5-year nonrelapse mortality (NRM) rate was not statistically different between the groups (13% in the younger group and 19% in the older group). By univariate and multivariate analysis, NRM was significantly higher in older patients who previously experienced failure with an autograft. The occurrence of grade 3 to 4 acute graft-versus-host disease (GVHD) or extensive chronic GVHD was associated with a higher NRM in both age cohorts. Overall survival (OS) was not statistically different between the younger (66%) and older groups (61%). By multivariate analysis, refractory disease was associated with a worse OS irrespective of age group.
CONCLUSION: RIC transplantations show a rather low NRM, and age 55 years per se cannot be considered a risk factor anymore. The timing of transplantation and novel strategies for the prevention of severe GVHD could further improve patient outcome.
INTRODUCTION
Allogeneic stem-cell transplantation (SCT) represents a potentially curative treatment for several hematologic malignancies, but myeloablative high-dose chemoradiotherapy used for conditioning is associated with a relevant incidence of organ toxicities that increase with advancing age and treatment lines. Analyses of transplantation registry data reported a rate of nonrelapse mortality (NRM) of up to 50% in patients older than 50 years.1-3 In particular, after a previously failed autograft, myeloablative conditioning was associated with an NRM rate ranging between 50% and 80%.4-6 Given the median age of patients with hematologic malignancies, the high NRM rates have limited the use of allogeneic SCT to a minority of young patients without serious comorbid medical conditions. Consequently, novel transplantation strategies have been explored during the last 5 years. Improvements in supportive care and a better selection of patients (ie, earlier transplantation and less prior chemotherapy) have encouraged some centers to treat patients more than 50 years old7-9; however, the notion that a relevant part of the curative potential of allogeneic SCT relies on the graft-versus-tumor activity10 prompted the investigation of less myelotoxic conditioning regimens. These regimens would not completely eradicate the tumor but would provide adequate immunosuppression for durable engraftment of donor stem cells.11,12
Several types of reduced-intensity conditioning (RIC) regimens provided a high rate of engraftment with a significant decrease in organ toxicity, making SCT a feasible procedure in patients considered not eligible for conventional transplantation because of age, comorbid conditions, or extensive previous therapies. Preliminary studies reported NRM rates ranging between 10% and 30%13-17 according to the conditioning regimen used and the patient characteristics (number of previous therapies, chemosensitive disease, and type of malignancy). We report the results of a large, prospective, multicenter, observational study investigating the impact of age and previously failed autograft on the NRM and survival of 150 consecutive patients affected by hematologic malignancies.
PATIENTS AND METHODS
Patient Characteristics
Between November 1998 and April 2004, 150 patients with advanced hematologic malignancies received a RIC regimen followed by allogeneic SCT from HLA-identical sibling donors in 10 Italian transplantation centers. Inclusion criteria were age 45 and 70 years and age younger than 45 years if the patient had serious comorbid conditions (cardiac ejection fraction < 50%, diffusion capacity of carbon monoxide < 50%, and abnormal liver function tests) or was heavily pretreated (> two lines of chemotherapy or a previously failed autologous transplantation). Approval was obtained from the institutional review board of each participating center, and all patients gave their written informed consent. Median age for the whole population was 52 years (range, 20 to 69 years). To evaluate the impact of age, patients were divided into two cohorts; 90 patients were younger than 55 years old, and 60 patients were 55 years old. The median age was 47 years for the younger group (range, 20 to 54 years) and 59 years for the older group (range, 55 to 69 years). Median follow-up time was 940 days for the younger group (range, 272 to 2,141 days) and 892 days for the older group (range, 172 to 1,760 days; P = .3). No statistical difference was found among all items evaluated in both groups (Table 1).
Study and Treatment Plan
This was a prospective, multicenter, observational phase II study. All patients received a RIC regimen consisting of thiotepa 10 mg/kg, cyclophosphamide 60 mg/kg, and fludarabine 60 mg/m2, followed by transplantation of marrow or granulocyte colony-stimulating factor mobilized blood hematopoietic stem cells on day 0.17 The CD34+ target dose was fixed and ranged between 4 and 8 x 106/kg recipient body weight when blood stem cells were used. Twenty patients received a bone marrow graft with a mononucleated cell target dose of greater than 3 x 108/kg recipient body weight.
Graft-versus-host disease (GVHD) prophylaxis consisted of cyclosporine (CSA) adjusted to 200 to 300 ng/mL blood levels and short-course methotrexate (10 mg/m2 on day +1 and 8 mg/m2 on days + 3 and +6). CSA was administered at full dose through day +100, and if GVHD did not occur, the dose was tapered by 10% every week thereafter. Patients with stable or progressive disease 40 to 60 days after transplantation rapidly received tapered CSA and then, if acute GVHD did not occur, received donor lymphocytes administered monthly in escalating doses, starting from 1 x 106 CD3/kg recipient body weight.
Statistical Methods
The characteristics of patients within each subgroup in the two age cohorts (Tables 1 and 2) were compared using the 2 test or the Fisher's exact test (two tailed). Data with a non-normal distribution for duration of follow-up were analyzed with a nonparametric test (Student's t test). The outcomes examined were overall survival (OS), NRM, and incidence of acute and chronic GVHD. Acute GVHD was classified as grade 1 to 4 according to Glucksberg et al,18 and chronic GVHD was classified as limited or extensive according to the criteria of Sullivan et al.19 OS included all causes of death and was estimated from time of transplantation to the date of death or last follow-up. NRM was defined as any death not caused by disease progression and was similarly estimated from time of transplantation to the date of death or last follow-up, except for patients who suffered disease progression whose survival experience was censored at the date of progression. All patients who were engrafted were considered assessable for acute GVHD, whereas 135 patients who survived more than 100 days were considered assessable for chronic GVHD. The closing date for analysis was December 31, 2004.
Estimates of OS and NRM were calculated according to the Kaplan-Meier method.20 The null hypothesis concerning the differential effects of putative prognostic factors in univariate analyses was tested by means of the log-rank test,21 and P values were two tailed. A multivariate analysis was performed using a Cox model.22 The null hypothesis of the regression analysis was tested by the Wald statistics.23 The relative risks were estimated as hazard ratios (HR).
RESULTS
Patient Outcome
All assessable patients had a sustained engraftment. The median time to recover an absolute neutrophil count of 0.5 x 109/L was 13 days (range, 9 to 22 days), and the median time to achieve platelets greater than 20 x 109/L was 15 days (range, 8 to 50 days). Only one patient died of toxicity before engraftment. As of December 2004, 102 of 150 patients are alive, with a median follow-up time of 927 days (range, 172 to 2,141 days). The estimated 5-year OS rate was 66% for younger patients and 61% for older patients (P = .25; Fig 1A). According to main pretransplantation prognostic factors, we found that, on univariate analysis, disease status and a previously failed autograft had a different influence on survival in both age groups. In particular, refractory disease was associated with a significantly worse survival among younger patients (P = .003), with a trend in the older group (P = .08); in contrast, a previously failed autograft was significantly associated with an inferior survival only in the older cohort (P = .002; Table 3).
Because of the apparent divergent effects on OS of refractory disease and previous autograft in the two age groups, the interactions between age and pretransplantation risk factors were assessed by multivariate analysis. Refractory disease was associated with a worse OS irrespective of age group (HR = 2.7; 95% CI, 1.49 to 4.9; P = .001). A previously failed autograft was associated with a worse OS (HR = 3.8; 95% CI, 1.57 to 9.24; P = .001) in the older age group only.
Acute myeloid leukemia and myelodysplasia patients were analyzed together, and the estimated 5-year OS and progression-free survival (PFS) rates were 32% and 38%, respectively. Multiple myeloma patients had estimated 5-year OS and PFS rates of 70% and 30%, respectively. Non-Hodgkin's lymphomas (NHL) were subdivided into indolent (follicular, mantle-cell, and small lymphocytic lymphomas) and aggressive (diffuse large B-cell and peripheral T-cell lymphomas). The 5-year OS and PFS rates for indolent NHL patients were 66% and 73%, respectively. Aggressive NHL patients had estimated 5-year OS and PFS rates of 72% and 59%, respectively (Figs 1B and 1C). Considering all patients together, a previously failed autograft had no influence on PFS (Fig 1D).
NRM
Eighteen (12%) of 150 patients died of NRM (eight patients in the younger group and 10 patients in the older group). The causes of death were infections (n = 10), acute GVHD (n = 4), congestive heart failure (n = 1), viral B hepatitis (n = 1), thrombotic thrombocytopenic purpura (n = 1), and primary graft failure (n = 1). The overall estimated 5-year NRM rate was 15% (13% for the younger group and 19% for the older group; P = .1; Fig 1E). No significant differences in NRM were seen splitting the whole population by disease categories (Fig 1F); moreover, when focusing the analysis on lymphoma patients, we did not find any significant difference in NRM between indolent and aggressive histologies both in the younger and older groups. On univariate analysis, a statistically significant association was found between a previously failed autograft and higher NRM in the older cohort; in fact, the estimated 5-year NRM rate was 11% for patients who did not experience failure of a previous autograft compared with 37% for patients who did experience a failure (P = .01; Fig 2A). Older patients with refractory disease had a significantly higher risk of NRM compared with chemosensitive patients (31% v 8%, respectively; P = .03; Fig 2B). In multivariate analysis, a statistically significant interaction was found between classes of age and both disease status (P = .03) and previous autografting (P = .02), indicating that older patients with refractory disease or a failed autograft had an NRM rate higher than the corresponding younger patients.
A significantly higher NRM was observed in patients developing grade 3 to 4 acute GVHD in both cohorts. In fact, in the younger group, the NRM rate was 24% for patients with grade 3 to 4 acute GVHD and 3% for patients with grade 1 to 2 acute GVHD (P = .01), whereas the older patients had NRM rates of 52% and 0%, respectively (P = .001; Figs 2C and 2D). When the effect of chronic GVHD on NRM was evaluated, we found a difference between limited and extensive chronic GVHD; the difference was statistically significant for younger patients (NRM rate, 0% v 28%, respectively; P = .04), whereas there was a trend in the older patients (NRM rate, 0% v 31%, respectively; P = .05; Figs 2E and 2F).
GVHD
All patients, except for one, survived after engraftment and were considered assessable for acute GVHD, whereas 135 patients surviving more than 100 days were considered also assessable for chronic GVHD. Overall, 89 patients (67%) developed signs of acute GVHD; 37 younger (41%) and 19 older patients (31%) had grade 1 to 2 acute GVHD, whereas 17 younger (28%) and 16 older patients (34%) developed grade 3 to 4 acute GVHD. Donor lymphocyte infusions were administered to 25 patients for persisting/progressive disease or for the treatment of relapse (10 patients in the younger cohort and 15 patients in the older cohort). As illustrated in Table 2, even when splitting the effect of donor lymphocyte infusions on the incidence of GVHD, no statistically significant difference was found between the two cohorts. There was no statistical difference also in the infection occurrence.
On univariate analysis, the onset of severe acute GVHD represented a significantly negative predictor for OS. The 5-year OS rates were 42% and 21% for young and older patients, respectively, with grade 3 to 4 acute GVHD compared with 5-year OS rates of 60% (P = .05) and 85% (P < .001) for younger and older patients, respectively, experiencing grade 1 to 2 acute GVHD (Figs 3A and 3B).
Chronic GVHD also had an impact on survival. In the younger group, OS rates were 82% and 45% for patients with limited and extensive chronic GVHD, respectively (P = .2); in the older group, the OS rates were 100% and 33% (P = .007) for patients with limited and extensive chronic GVHD, respectively (Figs 3C and 3D). Finally, in both cohorts, the onset of limited chronic GVHD provided a survival advantage compared with patients without chronic GVHD. In fact, 5-year OS for young patients with limited chronic GVHD was 82% v 60% for patients without any chronic GVHD (P = .05), and the corresponding OS rates were 100% v 53%, respectively, for the older group (P = .009; Figs 3E and 3F).
In the entire population of 150 patients, the influence of the degree of acute GVHD and pretransplantation risk factors was assessed by multivariate analysis in the two cohorts of patients, with grade 0 acute GVHD as the reference point. In the younger cohort, the development of grade 3 to 4 acute GVHD showed a significant impact on NRM (HR = 4.9; 95% CI, 1.21 to 19.7; P = .03). In the younger group, both grade 3 to 4 acute GVHD (HR = 2.47; 95% CI, 1.08 to 5.67; P = .03) and refractory disease (HR = 2.56; 95% CI, 1.09 to 5.98; P = .03) were associated with a statistically worse OS. In the older cohort, only grade 3 to 4 acute GVHD had a significantly negative influence (HR = 7.13; 95% CI, 1.82 to 27.82; P = .005) on NRM, whereas for OS, both grade 3 to 4 acute GVHD (HR = 5.27; 95% CI, 2.11 to 13.16; P = .003) and a previously failed autograft (HR = 4.36; 95% CI, 1.75 to 10.82; P = .001) had a significant impact.
DISCUSSION
In the present study, the impact of well-established transplantation risk factors (age, recipient sex, refractory disease, donor/recipient sex combination, number of lines of chemotherapy, previously failed autograft, and cytomegalovirus donor/recipient status) was analyzed on 150 patients affected by hematologic malignancies receiving the same RIC regimen followed by a match-related allogeneic SCT. Our findings indicate that NRM is quite low and compares favorably with the results obtained by other investigators in smaller studies.12-15,24
The main end point of our study was to analyze the impact of age on NRM by stratifying the population into two age cohorts. We chose to set the cutoff age at 55 years for two reasons. First, the onset of the majority of hematologic malignancies is usually after the fifth decade, and this population is growing steadily in Western countries. Second, it is traditionally considered the age limit for conventional allografting. Although several reports emphasized the feasibility of RIC transplantations in patients older than 50 to 55 years, there are few prospective data regarding the outcome of RIC when stratified for age.
The elderly are traditionally considered poor candidates for transplantation because they are usually affected by comorbidities and have received several treatment lines. When we performed our analysis by age, we observed a higher NRM rate in older patients, but it was not statistically significant. A recent multicenter Spanish study showed that age 60 years was associated with a significantly higher risk of NRM.25 A European Bone Marrow Transplantation Group retrospective study26 showed that age more than 50 years was a significant predictor for NRM; however, it was a registry study, not prospective, and included patients receiving several different RIC regimens, half of whom had an vivo T-cell depletion.
The second relevant end point of our study was to evaluate the impact of a previously failed autograft on NRM. Stratifying the entire population into two age cohorts, we found that NRM was significantly higher only in the older cohort, and the same results were observed when the analysis was also limited to lymphoma patients, which represented the largest disease category. Other recent studies have investigated the safety and efficacy of RIC transplantations after a failed autograft, giving conflicting results.25,27,28 However, the main limitation of these studies was that they considered the previously failed autograft as an independent risk factor and not related to age, so their results cannot be compared with ours.
Our findings provide a significant contribution to identifying a cohort of patients in whom the presence of an adverse prognostic factor, such as a previously failed autograft, increases significantly the toxicity of RIC transplantations. Consequently, the allogeneic SCT in this particular subset of patients should be taken into account more cautiously and used with proper timing during disease history.
Finally, the present study showed the critical role played by acute and chronic GVHD in the RIC setting. The incidence of GVHD depends on several factors including age, conditioning, donor type, and tissue damage induced by the preparative regimen. GVHD is reported in the literature to occur in 40% to 80% of myeloablative allograft recipients.29 Although it was first postulated that RIC regimens could lower the incidence of acute GVHD because of a reduction of tissue and mucosal barrier damage, at present, the incidence of acute GVHD after RIC remains a cause of concern, with a grade 3 to 4 GVHD rate of approximately 30%,13,14,30 which is a result similar to what is expected after a myeloablative conditioning.
Some studies on myeloablative transplantations have reported that acute GVHD incidence increases with age and postulated that this causes a higher NRM rate in older patients. In our series, older patients experienced severe acute GVHD more frequently. Acute GVHD had an impact on NRM; in fact, we observed a significant difference between NRM curves of patients experiencing moderate compared with severe acute GVHD. In particular, older patients seem to tolerate poorly the occurrence of grade 3 to 4 acute GVHD because 55% of them died of NRM. The same applied when the effect of chronic GVHD was evaluated; a significant difference was found in NRM of patients with limited or severe chronic GVHD. The results observed in univariate analysis were confirmed by multivariate analysis, in which we found that, irrespective of pretransplantation risk factors, acute or chronic GVHD retained a statistically significant impact on NRM. This is a relevant finding because we can postulate that novel strategies able to reduce the incidence of severe GVHD might impact NRM rates, particularly in the older group.
Our study was focused on the risk factors affecting NRM in the RIC setting, and a detailed analysis on disease outcomes was not performed. However, our OS and PFS results are encouraging and particularly positive for lymphoma patients and suggest that RIC transplantations can lower NRM while retaining a considerable antitumor activity.
Finally, considering the relatively long follow-up of our population, we have looked at the postulated immune effect of chronic GVHD on survival curves. In particular, we observed that actuarial 5-year OS rates in patients with limited chronic GVHD were projected to be 82% and 100% in the younger and older patients, respectively. These percentages were significantly higher, in both age cohorts, compared with patients not experiencing chronic GVHD.
In conclusion, this is the first study to show that age greater than 55 years is not a risk factor per se when a RIC regimen is used. Nevertheless, in the older age category, a previously failed autograft is a risk factor affecting NRM and OS. This implies that correct timing of allogeneic SCT can limit its toxicity. In addition, a major improvement in RIC programs should concern the reduction of severe acute GVHD.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the Italian Association for Cancer Research, and Compagnia di San Paolo "Programma Oncologia."
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Klingemann HG, Storb R, Fefer A, et al: Bone marrow transplantation in patients aged 45 years or older. Blood 67:770-776, 1986
Cahn JY, Labopin M, Schattenberg A, et al: Allogeneic bone marrow transplantation for acute leukemia in patients over the age of 40 years. Leukemia 11:416-419, 1997
Ringden O, Horowitz MM, Gale RP, et al: Outcome after allogeneic bone marrow transplant for leukemia in older adults. JAMA 270:57-60, 1993
Tsai T, Goodman S, Saez R, et al: Allogenic bone marrow transplantation in patients who relapse after autologous transplantation. Bone Marrow Transplant 20:859-863, 1997
Kulkarni S, Powles RL, Treleaven JG, et al: Impact of previous high-dose therapy on outcome after allografting for multiple myeloma. Bone Marrow Transplant 23:675-680, 1999
Radich JP, Gooley T, Sanders JE, et al: Second allogeneic transplantation after failure of first autologous transplantation. Biol Blood Marrow Transplant 6:272-279, 2000
Du W, Dansey R, Abella EM, et al: Successful allogeneic bone marrow transplantation in selected patients over 50 years of age: A single institution's experience. Bone Marrow Transplant 21:1043-1047, 1998
Deeg HJ, Shulman HM, Anderson JE, et al: Allogeneic and syngeneic marrow transplantation for myelodysplastic syndrome in patients 55 to 66 years of age. Blood 95:1188-1194, 2000
de la Camara R, Alonso A, Steegmann JL, et al: Allogeneic hematopoietic stem cell transplantation in patients 50 years of age and older. Haematologica 87:965-972, 2002
Giralt SA, Kolb HJ: Donor lymphocyte infusions. Curr Opin Oncol 8:96-102, 1996
Appelbaum FR: Haematopoietic cell transplantation as immunotherapy. Nature 411:385-389, 2001
Champlin R, Khouri I, Kornblau S, et al: Reinventing bone marrow transplantation: Reducing toxicity using nonmyeloablative, preparative regimens and induction of graft-versus-malignancy. Curr Opin Oncol 11:87-95, 1999
Slavin S, Nagler A, Naparstek E, et al: Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood 91:756-763, 1998
Giralt S, Estey E, Albitar M, et al: Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: Harnessing graft-versus-leukemia without myeloablative therapy. Blood 89:4531-4536, 1997
McSweeney PA, Niederwieser D, Shizuru JA, et al: Hematopoietic cell transplantation in older patients with hematologic malignancies: Replacing high-dose cytotoxic therapy with graft-versus-tumor effects. Blood 97:3390-3400, 2001
Kottaridis PD, Milligan DW, Chopra R, et al: In vivo CAMPATH-1H prevents graft-versus-host disease following nonmyeloablative stem cell transplantation. Blood 96:2419-2425, 2000
Corradini P, Tarella C, Olivieri A, et al: Reduced-intensity conditioning followed by allografting of hematopoietic cells can produce clinical and molecular remissions in patients with poor-risk hematologic malignancies. Blood 99:75-82, 2002
Glucksberg H, Storb R, Fefer A, et al: Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation 18:295-304, 1974
Sullivan KM, Shulman HM, Storb R, et al: Chronic graft-versus-host disease in 52 patients: Adverse natural course and successful treatment with combination immunosuppression. Blood 57:267-276, 1981
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Peto R, Peto J: Asymptotically efficient rank invariant test procedures. J R Stat Soc A 135:185-207, 1972
Cox Dr: Regression models and life-tables. J R Stat Soc B 34:187-220, 1972
Miller RG: Survival Analysis. New York, NY, John Wiley, 1981
Diaconescu R, Flowers CR, Storer B, et al: Morbidity and mortality with nonmyeloablative compared with myeloablative conditioning before hematopoietic cell transplantation from HLA-matched related donors. Blood 104:1550-1558, 2004
Gomez-Nunez M, Martino R, Caballero MD, et al: Elderly age and prior autologous transplantation have a deleterious effect on survival following allogeneic peripheral blood stem cell transplantation with reduced-intensity conditioning: Results from the Spanish multicenter prospective trial. Bone Marrow Transplant 33:477-482, 2004
Robinson SP, Goldstone AH, Mackinnon S, et al: Chemoresistant or aggressive lymphoma predicts for a poor outcome following reduced-intensity allogeneic progenitor cell transplantation: An analysis from the Lymphoma Working Party of the European Group for Blood and Bone Marrow Transplantation. Blood 100:4310-4316, 2002
Branson K, Chopra R, Kottaridis PD, et al: Role of nonmyeloablative allogeneic stem-cell transplantation after failure of autologous transplantation in patients with lymphoproliferative malignancies. J Clin Oncol 20:4022-4031, 2002
Escalon MP, Champlin RE, Saliba RM, et al: Nonmyeloablative allogeneic hematopoietic transplantation: A promising salvage therapy for patients with non-Hodgkin's lymphoma whose disease has failed a prior autologous transplantation. J Clin Oncol 22:2419-2423, 2004
Marcellus DC, Vogelsang GB: Graft-versus-host disease. Curr Opin Oncol 9:131-138, 1997
Schetelig J, Kroger N, Held TK, et al: Allogeneic transplantation after reduced conditioning in high risk patients is complicated by a high incidence of acute and chronic graft-versus-host disease. Haematologica 87:299-305, 2002(P. Corradini, F. Zallio, )
Divisione di Ematologia, Unità Trapianto Midollo, Ospedale San Raffaele, Milano
Divisione di Ematologia II, Ospedale San Martino, Genova
Divisione di Ematologia, Università di Ancona, Ancona
Divisione di Ematologia, Università di Udine, Udine
Divisione di Ematologia, Ospedale Riuniti, Bergamo
Unità Trapianto Midollo, Ospedale Cervello, Palermo
Divisione di Ematologia, Ospedale San Camillo-Forlanini, Roma
Istituto di Ematologia–Oncologia, Università di Bologna, Bologna
Divisione di Ematologia, Università di Torino, Torino, Italy
ABSTRACT
PURPOSE: Older age and a previously failed autologous stem-cell transplantation (SCT) are poor prognostic factors for patients receiving myeloablative conditioning and allogeneic SCT. Reduced-intensity conditioning (RIC) regimens achieved a significant reduction of treatment-related mortality, but the influence of previously described risk factors on the outcome of this novel transplantation strategy have not been fully analyzed yet.
PATIENTS AND METHODS: One hundred fifty patients with advanced hematologic malignancies received a RIC regimen containing thiotepa (10 mg/kg), fludarabine (60 mg/m2), and cyclophosphamide (60 mg/kg), followed by an allogeneic transplantation from an HLA-identical sibling donor. Patients were divided into two cohorts according to age; 90 patients were younger than 55 years, and 60 patients were 55 years old or older. The other pretransplantation characteristics were fairly balanced.
RESULTS: Actuarial 5-year nonrelapse mortality (NRM) rate was not statistically different between the groups (13% in the younger group and 19% in the older group). By univariate and multivariate analysis, NRM was significantly higher in older patients who previously experienced failure with an autograft. The occurrence of grade 3 to 4 acute graft-versus-host disease (GVHD) or extensive chronic GVHD was associated with a higher NRM in both age cohorts. Overall survival (OS) was not statistically different between the younger (66%) and older groups (61%). By multivariate analysis, refractory disease was associated with a worse OS irrespective of age group.
CONCLUSION: RIC transplantations show a rather low NRM, and age 55 years per se cannot be considered a risk factor anymore. The timing of transplantation and novel strategies for the prevention of severe GVHD could further improve patient outcome.
INTRODUCTION
Allogeneic stem-cell transplantation (SCT) represents a potentially curative treatment for several hematologic malignancies, but myeloablative high-dose chemoradiotherapy used for conditioning is associated with a relevant incidence of organ toxicities that increase with advancing age and treatment lines. Analyses of transplantation registry data reported a rate of nonrelapse mortality (NRM) of up to 50% in patients older than 50 years.1-3 In particular, after a previously failed autograft, myeloablative conditioning was associated with an NRM rate ranging between 50% and 80%.4-6 Given the median age of patients with hematologic malignancies, the high NRM rates have limited the use of allogeneic SCT to a minority of young patients without serious comorbid medical conditions. Consequently, novel transplantation strategies have been explored during the last 5 years. Improvements in supportive care and a better selection of patients (ie, earlier transplantation and less prior chemotherapy) have encouraged some centers to treat patients more than 50 years old7-9; however, the notion that a relevant part of the curative potential of allogeneic SCT relies on the graft-versus-tumor activity10 prompted the investigation of less myelotoxic conditioning regimens. These regimens would not completely eradicate the tumor but would provide adequate immunosuppression for durable engraftment of donor stem cells.11,12
Several types of reduced-intensity conditioning (RIC) regimens provided a high rate of engraftment with a significant decrease in organ toxicity, making SCT a feasible procedure in patients considered not eligible for conventional transplantation because of age, comorbid conditions, or extensive previous therapies. Preliminary studies reported NRM rates ranging between 10% and 30%13-17 according to the conditioning regimen used and the patient characteristics (number of previous therapies, chemosensitive disease, and type of malignancy). We report the results of a large, prospective, multicenter, observational study investigating the impact of age and previously failed autograft on the NRM and survival of 150 consecutive patients affected by hematologic malignancies.
PATIENTS AND METHODS
Patient Characteristics
Between November 1998 and April 2004, 150 patients with advanced hematologic malignancies received a RIC regimen followed by allogeneic SCT from HLA-identical sibling donors in 10 Italian transplantation centers. Inclusion criteria were age 45 and 70 years and age younger than 45 years if the patient had serious comorbid conditions (cardiac ejection fraction < 50%, diffusion capacity of carbon monoxide < 50%, and abnormal liver function tests) or was heavily pretreated (> two lines of chemotherapy or a previously failed autologous transplantation). Approval was obtained from the institutional review board of each participating center, and all patients gave their written informed consent. Median age for the whole population was 52 years (range, 20 to 69 years). To evaluate the impact of age, patients were divided into two cohorts; 90 patients were younger than 55 years old, and 60 patients were 55 years old. The median age was 47 years for the younger group (range, 20 to 54 years) and 59 years for the older group (range, 55 to 69 years). Median follow-up time was 940 days for the younger group (range, 272 to 2,141 days) and 892 days for the older group (range, 172 to 1,760 days; P = .3). No statistical difference was found among all items evaluated in both groups (Table 1).
Study and Treatment Plan
This was a prospective, multicenter, observational phase II study. All patients received a RIC regimen consisting of thiotepa 10 mg/kg, cyclophosphamide 60 mg/kg, and fludarabine 60 mg/m2, followed by transplantation of marrow or granulocyte colony-stimulating factor mobilized blood hematopoietic stem cells on day 0.17 The CD34+ target dose was fixed and ranged between 4 and 8 x 106/kg recipient body weight when blood stem cells were used. Twenty patients received a bone marrow graft with a mononucleated cell target dose of greater than 3 x 108/kg recipient body weight.
Graft-versus-host disease (GVHD) prophylaxis consisted of cyclosporine (CSA) adjusted to 200 to 300 ng/mL blood levels and short-course methotrexate (10 mg/m2 on day +1 and 8 mg/m2 on days + 3 and +6). CSA was administered at full dose through day +100, and if GVHD did not occur, the dose was tapered by 10% every week thereafter. Patients with stable or progressive disease 40 to 60 days after transplantation rapidly received tapered CSA and then, if acute GVHD did not occur, received donor lymphocytes administered monthly in escalating doses, starting from 1 x 106 CD3/kg recipient body weight.
Statistical Methods
The characteristics of patients within each subgroup in the two age cohorts (Tables 1 and 2) were compared using the 2 test or the Fisher's exact test (two tailed). Data with a non-normal distribution for duration of follow-up were analyzed with a nonparametric test (Student's t test). The outcomes examined were overall survival (OS), NRM, and incidence of acute and chronic GVHD. Acute GVHD was classified as grade 1 to 4 according to Glucksberg et al,18 and chronic GVHD was classified as limited or extensive according to the criteria of Sullivan et al.19 OS included all causes of death and was estimated from time of transplantation to the date of death or last follow-up. NRM was defined as any death not caused by disease progression and was similarly estimated from time of transplantation to the date of death or last follow-up, except for patients who suffered disease progression whose survival experience was censored at the date of progression. All patients who were engrafted were considered assessable for acute GVHD, whereas 135 patients who survived more than 100 days were considered assessable for chronic GVHD. The closing date for analysis was December 31, 2004.
Estimates of OS and NRM were calculated according to the Kaplan-Meier method.20 The null hypothesis concerning the differential effects of putative prognostic factors in univariate analyses was tested by means of the log-rank test,21 and P values were two tailed. A multivariate analysis was performed using a Cox model.22 The null hypothesis of the regression analysis was tested by the Wald statistics.23 The relative risks were estimated as hazard ratios (HR).
RESULTS
Patient Outcome
All assessable patients had a sustained engraftment. The median time to recover an absolute neutrophil count of 0.5 x 109/L was 13 days (range, 9 to 22 days), and the median time to achieve platelets greater than 20 x 109/L was 15 days (range, 8 to 50 days). Only one patient died of toxicity before engraftment. As of December 2004, 102 of 150 patients are alive, with a median follow-up time of 927 days (range, 172 to 2,141 days). The estimated 5-year OS rate was 66% for younger patients and 61% for older patients (P = .25; Fig 1A). According to main pretransplantation prognostic factors, we found that, on univariate analysis, disease status and a previously failed autograft had a different influence on survival in both age groups. In particular, refractory disease was associated with a significantly worse survival among younger patients (P = .003), with a trend in the older group (P = .08); in contrast, a previously failed autograft was significantly associated with an inferior survival only in the older cohort (P = .002; Table 3).
Because of the apparent divergent effects on OS of refractory disease and previous autograft in the two age groups, the interactions between age and pretransplantation risk factors were assessed by multivariate analysis. Refractory disease was associated with a worse OS irrespective of age group (HR = 2.7; 95% CI, 1.49 to 4.9; P = .001). A previously failed autograft was associated with a worse OS (HR = 3.8; 95% CI, 1.57 to 9.24; P = .001) in the older age group only.
Acute myeloid leukemia and myelodysplasia patients were analyzed together, and the estimated 5-year OS and progression-free survival (PFS) rates were 32% and 38%, respectively. Multiple myeloma patients had estimated 5-year OS and PFS rates of 70% and 30%, respectively. Non-Hodgkin's lymphomas (NHL) were subdivided into indolent (follicular, mantle-cell, and small lymphocytic lymphomas) and aggressive (diffuse large B-cell and peripheral T-cell lymphomas). The 5-year OS and PFS rates for indolent NHL patients were 66% and 73%, respectively. Aggressive NHL patients had estimated 5-year OS and PFS rates of 72% and 59%, respectively (Figs 1B and 1C). Considering all patients together, a previously failed autograft had no influence on PFS (Fig 1D).
NRM
Eighteen (12%) of 150 patients died of NRM (eight patients in the younger group and 10 patients in the older group). The causes of death were infections (n = 10), acute GVHD (n = 4), congestive heart failure (n = 1), viral B hepatitis (n = 1), thrombotic thrombocytopenic purpura (n = 1), and primary graft failure (n = 1). The overall estimated 5-year NRM rate was 15% (13% for the younger group and 19% for the older group; P = .1; Fig 1E). No significant differences in NRM were seen splitting the whole population by disease categories (Fig 1F); moreover, when focusing the analysis on lymphoma patients, we did not find any significant difference in NRM between indolent and aggressive histologies both in the younger and older groups. On univariate analysis, a statistically significant association was found between a previously failed autograft and higher NRM in the older cohort; in fact, the estimated 5-year NRM rate was 11% for patients who did not experience failure of a previous autograft compared with 37% for patients who did experience a failure (P = .01; Fig 2A). Older patients with refractory disease had a significantly higher risk of NRM compared with chemosensitive patients (31% v 8%, respectively; P = .03; Fig 2B). In multivariate analysis, a statistically significant interaction was found between classes of age and both disease status (P = .03) and previous autografting (P = .02), indicating that older patients with refractory disease or a failed autograft had an NRM rate higher than the corresponding younger patients.
A significantly higher NRM was observed in patients developing grade 3 to 4 acute GVHD in both cohorts. In fact, in the younger group, the NRM rate was 24% for patients with grade 3 to 4 acute GVHD and 3% for patients with grade 1 to 2 acute GVHD (P = .01), whereas the older patients had NRM rates of 52% and 0%, respectively (P = .001; Figs 2C and 2D). When the effect of chronic GVHD on NRM was evaluated, we found a difference between limited and extensive chronic GVHD; the difference was statistically significant for younger patients (NRM rate, 0% v 28%, respectively; P = .04), whereas there was a trend in the older patients (NRM rate, 0% v 31%, respectively; P = .05; Figs 2E and 2F).
GVHD
All patients, except for one, survived after engraftment and were considered assessable for acute GVHD, whereas 135 patients surviving more than 100 days were considered also assessable for chronic GVHD. Overall, 89 patients (67%) developed signs of acute GVHD; 37 younger (41%) and 19 older patients (31%) had grade 1 to 2 acute GVHD, whereas 17 younger (28%) and 16 older patients (34%) developed grade 3 to 4 acute GVHD. Donor lymphocyte infusions were administered to 25 patients for persisting/progressive disease or for the treatment of relapse (10 patients in the younger cohort and 15 patients in the older cohort). As illustrated in Table 2, even when splitting the effect of donor lymphocyte infusions on the incidence of GVHD, no statistically significant difference was found between the two cohorts. There was no statistical difference also in the infection occurrence.
On univariate analysis, the onset of severe acute GVHD represented a significantly negative predictor for OS. The 5-year OS rates were 42% and 21% for young and older patients, respectively, with grade 3 to 4 acute GVHD compared with 5-year OS rates of 60% (P = .05) and 85% (P < .001) for younger and older patients, respectively, experiencing grade 1 to 2 acute GVHD (Figs 3A and 3B).
Chronic GVHD also had an impact on survival. In the younger group, OS rates were 82% and 45% for patients with limited and extensive chronic GVHD, respectively (P = .2); in the older group, the OS rates were 100% and 33% (P = .007) for patients with limited and extensive chronic GVHD, respectively (Figs 3C and 3D). Finally, in both cohorts, the onset of limited chronic GVHD provided a survival advantage compared with patients without chronic GVHD. In fact, 5-year OS for young patients with limited chronic GVHD was 82% v 60% for patients without any chronic GVHD (P = .05), and the corresponding OS rates were 100% v 53%, respectively, for the older group (P = .009; Figs 3E and 3F).
In the entire population of 150 patients, the influence of the degree of acute GVHD and pretransplantation risk factors was assessed by multivariate analysis in the two cohorts of patients, with grade 0 acute GVHD as the reference point. In the younger cohort, the development of grade 3 to 4 acute GVHD showed a significant impact on NRM (HR = 4.9; 95% CI, 1.21 to 19.7; P = .03). In the younger group, both grade 3 to 4 acute GVHD (HR = 2.47; 95% CI, 1.08 to 5.67; P = .03) and refractory disease (HR = 2.56; 95% CI, 1.09 to 5.98; P = .03) were associated with a statistically worse OS. In the older cohort, only grade 3 to 4 acute GVHD had a significantly negative influence (HR = 7.13; 95% CI, 1.82 to 27.82; P = .005) on NRM, whereas for OS, both grade 3 to 4 acute GVHD (HR = 5.27; 95% CI, 2.11 to 13.16; P = .003) and a previously failed autograft (HR = 4.36; 95% CI, 1.75 to 10.82; P = .001) had a significant impact.
DISCUSSION
In the present study, the impact of well-established transplantation risk factors (age, recipient sex, refractory disease, donor/recipient sex combination, number of lines of chemotherapy, previously failed autograft, and cytomegalovirus donor/recipient status) was analyzed on 150 patients affected by hematologic malignancies receiving the same RIC regimen followed by a match-related allogeneic SCT. Our findings indicate that NRM is quite low and compares favorably with the results obtained by other investigators in smaller studies.12-15,24
The main end point of our study was to analyze the impact of age on NRM by stratifying the population into two age cohorts. We chose to set the cutoff age at 55 years for two reasons. First, the onset of the majority of hematologic malignancies is usually after the fifth decade, and this population is growing steadily in Western countries. Second, it is traditionally considered the age limit for conventional allografting. Although several reports emphasized the feasibility of RIC transplantations in patients older than 50 to 55 years, there are few prospective data regarding the outcome of RIC when stratified for age.
The elderly are traditionally considered poor candidates for transplantation because they are usually affected by comorbidities and have received several treatment lines. When we performed our analysis by age, we observed a higher NRM rate in older patients, but it was not statistically significant. A recent multicenter Spanish study showed that age 60 years was associated with a significantly higher risk of NRM.25 A European Bone Marrow Transplantation Group retrospective study26 showed that age more than 50 years was a significant predictor for NRM; however, it was a registry study, not prospective, and included patients receiving several different RIC regimens, half of whom had an vivo T-cell depletion.
The second relevant end point of our study was to evaluate the impact of a previously failed autograft on NRM. Stratifying the entire population into two age cohorts, we found that NRM was significantly higher only in the older cohort, and the same results were observed when the analysis was also limited to lymphoma patients, which represented the largest disease category. Other recent studies have investigated the safety and efficacy of RIC transplantations after a failed autograft, giving conflicting results.25,27,28 However, the main limitation of these studies was that they considered the previously failed autograft as an independent risk factor and not related to age, so their results cannot be compared with ours.
Our findings provide a significant contribution to identifying a cohort of patients in whom the presence of an adverse prognostic factor, such as a previously failed autograft, increases significantly the toxicity of RIC transplantations. Consequently, the allogeneic SCT in this particular subset of patients should be taken into account more cautiously and used with proper timing during disease history.
Finally, the present study showed the critical role played by acute and chronic GVHD in the RIC setting. The incidence of GVHD depends on several factors including age, conditioning, donor type, and tissue damage induced by the preparative regimen. GVHD is reported in the literature to occur in 40% to 80% of myeloablative allograft recipients.29 Although it was first postulated that RIC regimens could lower the incidence of acute GVHD because of a reduction of tissue and mucosal barrier damage, at present, the incidence of acute GVHD after RIC remains a cause of concern, with a grade 3 to 4 GVHD rate of approximately 30%,13,14,30 which is a result similar to what is expected after a myeloablative conditioning.
Some studies on myeloablative transplantations have reported that acute GVHD incidence increases with age and postulated that this causes a higher NRM rate in older patients. In our series, older patients experienced severe acute GVHD more frequently. Acute GVHD had an impact on NRM; in fact, we observed a significant difference between NRM curves of patients experiencing moderate compared with severe acute GVHD. In particular, older patients seem to tolerate poorly the occurrence of grade 3 to 4 acute GVHD because 55% of them died of NRM. The same applied when the effect of chronic GVHD was evaluated; a significant difference was found in NRM of patients with limited or severe chronic GVHD. The results observed in univariate analysis were confirmed by multivariate analysis, in which we found that, irrespective of pretransplantation risk factors, acute or chronic GVHD retained a statistically significant impact on NRM. This is a relevant finding because we can postulate that novel strategies able to reduce the incidence of severe GVHD might impact NRM rates, particularly in the older group.
Our study was focused on the risk factors affecting NRM in the RIC setting, and a detailed analysis on disease outcomes was not performed. However, our OS and PFS results are encouraging and particularly positive for lymphoma patients and suggest that RIC transplantations can lower NRM while retaining a considerable antitumor activity.
Finally, considering the relatively long follow-up of our population, we have looked at the postulated immune effect of chronic GVHD on survival curves. In particular, we observed that actuarial 5-year OS rates in patients with limited chronic GVHD were projected to be 82% and 100% in the younger and older patients, respectively. These percentages were significantly higher, in both age cohorts, compared with patients not experiencing chronic GVHD.
In conclusion, this is the first study to show that age greater than 55 years is not a risk factor per se when a RIC regimen is used. Nevertheless, in the older age category, a previously failed autograft is a risk factor affecting NRM and OS. This implies that correct timing of allogeneic SCT can limit its toxicity. In addition, a major improvement in RIC programs should concern the reduction of severe acute GVHD.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by the Italian Association for Cancer Research, and Compagnia di San Paolo "Programma Oncologia."
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Klingemann HG, Storb R, Fefer A, et al: Bone marrow transplantation in patients aged 45 years or older. Blood 67:770-776, 1986
Cahn JY, Labopin M, Schattenberg A, et al: Allogeneic bone marrow transplantation for acute leukemia in patients over the age of 40 years. Leukemia 11:416-419, 1997
Ringden O, Horowitz MM, Gale RP, et al: Outcome after allogeneic bone marrow transplant for leukemia in older adults. JAMA 270:57-60, 1993
Tsai T, Goodman S, Saez R, et al: Allogenic bone marrow transplantation in patients who relapse after autologous transplantation. Bone Marrow Transplant 20:859-863, 1997
Kulkarni S, Powles RL, Treleaven JG, et al: Impact of previous high-dose therapy on outcome after allografting for multiple myeloma. Bone Marrow Transplant 23:675-680, 1999
Radich JP, Gooley T, Sanders JE, et al: Second allogeneic transplantation after failure of first autologous transplantation. Biol Blood Marrow Transplant 6:272-279, 2000
Du W, Dansey R, Abella EM, et al: Successful allogeneic bone marrow transplantation in selected patients over 50 years of age: A single institution's experience. Bone Marrow Transplant 21:1043-1047, 1998
Deeg HJ, Shulman HM, Anderson JE, et al: Allogeneic and syngeneic marrow transplantation for myelodysplastic syndrome in patients 55 to 66 years of age. Blood 95:1188-1194, 2000
de la Camara R, Alonso A, Steegmann JL, et al: Allogeneic hematopoietic stem cell transplantation in patients 50 years of age and older. Haematologica 87:965-972, 2002
Giralt SA, Kolb HJ: Donor lymphocyte infusions. Curr Opin Oncol 8:96-102, 1996
Appelbaum FR: Haematopoietic cell transplantation as immunotherapy. Nature 411:385-389, 2001
Champlin R, Khouri I, Kornblau S, et al: Reinventing bone marrow transplantation: Reducing toxicity using nonmyeloablative, preparative regimens and induction of graft-versus-malignancy. Curr Opin Oncol 11:87-95, 1999
Slavin S, Nagler A, Naparstek E, et al: Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood 91:756-763, 1998
Giralt S, Estey E, Albitar M, et al: Engraftment of allogeneic hematopoietic progenitor cells with purine analog-containing chemotherapy: Harnessing graft-versus-leukemia without myeloablative therapy. Blood 89:4531-4536, 1997
McSweeney PA, Niederwieser D, Shizuru JA, et al: Hematopoietic cell transplantation in older patients with hematologic malignancies: Replacing high-dose cytotoxic therapy with graft-versus-tumor effects. Blood 97:3390-3400, 2001
Kottaridis PD, Milligan DW, Chopra R, et al: In vivo CAMPATH-1H prevents graft-versus-host disease following nonmyeloablative stem cell transplantation. Blood 96:2419-2425, 2000
Corradini P, Tarella C, Olivieri A, et al: Reduced-intensity conditioning followed by allografting of hematopoietic cells can produce clinical and molecular remissions in patients with poor-risk hematologic malignancies. Blood 99:75-82, 2002
Glucksberg H, Storb R, Fefer A, et al: Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation 18:295-304, 1974
Sullivan KM, Shulman HM, Storb R, et al: Chronic graft-versus-host disease in 52 patients: Adverse natural course and successful treatment with combination immunosuppression. Blood 57:267-276, 1981
Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457-481, 1958
Peto R, Peto J: Asymptotically efficient rank invariant test procedures. J R Stat Soc A 135:185-207, 1972
Cox Dr: Regression models and life-tables. J R Stat Soc B 34:187-220, 1972
Miller RG: Survival Analysis. New York, NY, John Wiley, 1981
Diaconescu R, Flowers CR, Storer B, et al: Morbidity and mortality with nonmyeloablative compared with myeloablative conditioning before hematopoietic cell transplantation from HLA-matched related donors. Blood 104:1550-1558, 2004
Gomez-Nunez M, Martino R, Caballero MD, et al: Elderly age and prior autologous transplantation have a deleterious effect on survival following allogeneic peripheral blood stem cell transplantation with reduced-intensity conditioning: Results from the Spanish multicenter prospective trial. Bone Marrow Transplant 33:477-482, 2004
Robinson SP, Goldstone AH, Mackinnon S, et al: Chemoresistant or aggressive lymphoma predicts for a poor outcome following reduced-intensity allogeneic progenitor cell transplantation: An analysis from the Lymphoma Working Party of the European Group for Blood and Bone Marrow Transplantation. Blood 100:4310-4316, 2002
Branson K, Chopra R, Kottaridis PD, et al: Role of nonmyeloablative allogeneic stem-cell transplantation after failure of autologous transplantation in patients with lymphoproliferative malignancies. J Clin Oncol 20:4022-4031, 2002
Escalon MP, Champlin RE, Saliba RM, et al: Nonmyeloablative allogeneic hematopoietic transplantation: A promising salvage therapy for patients with non-Hodgkin's lymphoma whose disease has failed a prior autologous transplantation. J Clin Oncol 22:2419-2423, 2004
Marcellus DC, Vogelsang GB: Graft-versus-host disease. Curr Opin Oncol 9:131-138, 1997
Schetelig J, Kroger N, Held TK, et al: Allogeneic transplantation after reduced conditioning in high risk patients is complicated by a high incidence of acute and chronic graft-versus-host disease. Haematologica 87:299-305, 2002(P. Corradini, F. Zallio, )