Stem cell and cord blood transplantation – state of the art


  • 6C-PL7-03

Priv.-Doz. Dr. med. Gesine Bug, Medizinische Klinik 2, Klinikum der Johann-Wolfgang-Goethe-Universität, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany


Recent advances have broadened the application of allogeneic hematopoietic stem cell (HSC) transplantation and contributed to the continuously increasing numbers of transplantations performed worldwide. These include (1) greater utilization of reduced intensity conditioning and improvement of supportive care allowing transplantation of patients up to 70 years of age and with pre-existing medical problems, and (2) expansion of the acceptable stem cell donor pool to unrelated cord blood HSC. Thus, selection of the particular transplant procedure should be guided by patient characteristics such as type and stage of the disease, previous therapies, age and comorbidities. HLA typing of patient and siblings at diagnosis is essential to allow the timely initiation of an unrelated donor search if needed.


An allogeneic hematopoietic stem cell transplantation (HSCT) provides the best, and sometimes only, chance of long-term survival for many patients with otherwise fatal hematologic diseases such as acute leukaemias. The curative potential of HSCT is partly attributable to a graft-versus-leukaemia effect generated by reactivity of donor cells against recipient tissues, including residual leukaemia cells [1]. Thus, HSCT can be conceived as the first cancer immune therapy and the earliest example of an individualized cancer treatment [2]. In detail, the transplanting physician will have to consider (1) the stem cell donor and source; (2) the intensity and composition of the conditioning regimen including total body irradiation (TBI) and/or chemotherapy to reduce the tumour burden; (3) immunosuppressive therapy to avoid graft rejection and graft-versus-host disease; (4) prevention and treatment of infections emerging immediately post-transplant during neutropenia or later until recovery of T-cell function. Here, we present state of the art information on HSCT in adult patients with acute leukaemia.

Stem cell donors

Donor types for HSCT are categorized as HLA-identical sibling, other family donor or unrelated donor. According to a German report, a matched sibling is found in 40% of searches, and a one locus mismatched sibling or another relative in 3·1% and 10·4% of cases, respectively [3]. A matched unrelated donor (MUD) is defined as a 9/10 or 10/10 allelically identical donor based on HLA high-resolution typing for class I (HLA-A, -B, -C) and class II (HLA-DRB1, -DQB1). In a prospective, randomized trial including 236 patients with myeloid malignancies, HSCT from 10/10 MUD led to outcomes similar to those from HLA-identical sibling donors. In contrast, transplant-related morbidity and mortality resulting from graft rejection as well as acute and chronic graft-versus-host disease (GVHD) are significantly increased with a mismatched donor [4,5].

As of March 2010, more than 13 000 000 voluntary donors have been registered worldwide [6]. An unrelated donor search yields an acceptable match in approximately 75% of cases, but it can require 3–4 months, and about one-third of donors are not available at the time they are needed [7]. In addition, 400 000 cord blood units have been collected during the last 10 years and are an alternative stem cell source for patients with acute leukaemias in need of immediate HSCT who do not find a MUD. In 2007, umbilical cord blood transplantation (UCBT) comprised 28% of all unrelated procedures performed [8].

Principles of conditioning

A successful HSCT for a patient with a hematologic malignancy depends on providing adequate cytotoxic and immunosuppressive therapy. Until recently, most patients with acute leukaemias were given the same myeloablative conditioning (MAC) regimen consisting of TBI or busulfan (Bu) for eradication of the normal as well as malignant hematopoietic system and cyclophosphamide (Cy) for immunosuppression. TBI/Cy and Bu/Cy produce similar long-term results in acute myeloid leukaemia (AML) patients, whereas TBI-containing regimens are preferred for patients with acute lymphoblastic leukaemia (ALL). Because of their high non-hematologic toxicity, MAC regimens have primarily been delivered to younger recipients up to the age of 50 or 55 , excluding the majority of patients with acute leukaemia.

Non-myeloablative or reduced intensity conditioning (RIC) regimens take into account the particular risk of the disease and of treatment-related mortality (TRM) for individual patients. They are based on the concept of HSCT as an immune therapy and aim to achieve engraftment of donor stem cells with use of minimal doses of radiation and/or chemotherapy. After transplantation, donor lymphocyte infusions (DLI) can be given to convert incomplete donor chimerism. In patients with AML, RIC HSCT is associated with lower early organ toxicity, but impact on GVHD and infectious complications is much less, and relapse rate is always increased [9].

Indications for HSCT

According to a survey of the European Group of Blood and Bone Marrow Transplantation (EBMT) on the European HSCT activity in 2007, the main indications for an allogeneic HSCT were leukaemias (7174 patients), lymphoproliferative diseases (1608 patients) and non-malignant diseases (1184 patients) [10].

In patients with acute leukaemia, the decision for an HSCT depends on the risk of the disease. Thus, a complete work-up of genetic, immunophenotypic and clinical risk factors is mandatory for each patient with newly diagnosed acute myeloid or lymphoblastic leukaemia and is complemented by early evaluation of the diseases’ response to standard induction chemotherapy.

HSCT for acute myeloid leukaemia

High risk patients are defined by (1) unfavourable risk cytogenetics or distinctive mutations such as Flt3 internal tandem duplications and/or (2) delayed response to standard induction therapy even if they finally achieve complete remission (CR) and/or (3) relapse at any time during or after chemotherapy. Patients in first CR without favourable cytogenetics including t(8;21) or inv(16) or genetically defined high risk features fall into the standard risk group [11,12]. Depending on the particular risk, HSCT may be used as planned consolidation in first CR, as salvage therapy for patients refractory to standard induction chemotherapy, at first relapse, or in second CR [13].

AML in first complete remission

For adult standard risk patients aged <60  with a matched related donor, the appropriate consolidation therapy has been addressed by numerous prospective trials. Typically, patients were enrolled at diagnosis, and those who achieved CR were genetically randomized, i.e. patients were assigned to HSCT if they had a HLA-matched sibling. Consolidation of patients without a donor consisted of chemotherapy or autografting. In spite of large patient numbers included, the intent-to-treat analyses delivered inconsistent data, and the main confounders are associated with disease recurrence: In the donor group, a drop-out rate of up to 40% of patients, who did not undergo HSCT mainly because of early relapse, was observed. In the no donor group, patients relapsing during follow-up had HSCT as salvage therapy. In contrast, an excellent adherence to the protocol resulting in an HSCT rate of >80% in the donor group has been achieved by the Dutch-Belgian-Swiss HOVON/SAKK study group [14] and resulted in a significant benefit in overall survival by donor availability for all patients with AML in first CR without favourable cytogenetics. At 4 years, 53% of the actually transplanted patients were still in first remission, 22% had relapsed and 24% died of transplant-related reasons.

The authors also performed a meta-analysis based on data of more than 4000 patients treated in protocols of large European study groups and confirmed a significant overall survival benefit of 12% for patients with standard and high risk AML. They concluded that the survival benefit became evident as soon as the relapse risk exceeded 35%. In patients older than 35 years, the survival benefit was less pronounced, because of a higher TRM rate. Finally, the most comprehensive meta-analysis [15] including over 6000 patients with AML in first CR treated between 1982 and 2006 also supports a risk-based approach to offer HSCT to all patients with standard and high risk AML with an HLA-identical sibling under the age of 55 or 60.

The only prospective trial [16] evaluating the impact of MUD HSCT for AML in first CR has been published in abstract form only. Patients classified as high risk were allocated to undergo either HSCT (n = 34 from a matched sibling donor and n = 29 from a MUD) or autografting (n = 26). Overall survival at 4 years was significantly better in the HSCT recipients (68% and 56%, respectively) compared to the autografted patients (23%). Given additional, uncontrolled trials, which report superior outcomes in HSCT recipients with high risk AML, irrespective of a matched related or unrelated donor has been used [17], HSCT should be aimed for each patient with AML with unfavourable genetic features.

Relapsed and refractory AML

Patients with AML who relapse within 6 months after diagnosis are considered to have refractory disease. They often do not achieve a second remission upon conventional chemotherapy and have a very poor prognosis. Recently, HSCT protocols consisting of a first block of intensive chemotherapy and – after an interval of at least 3 days – a second block of a RIC regimen have been successfully applied to patients with refractory AML and resulted in overall survival rates of 40% at 2 years [18]. Of note, most of the 103 patients enrolled had active disease. Even better overall survival rates of 72–75% at 2 years have been achieved if high risk patients were transplanted in first CR [19] or in aplasia after standard induction chemotherapy [20]. Another approach combines a high-dose anti-leukaemic and immunosuppressive treatment, leading to similar overall survival rates [21]. In conclusion, these protocols deliver an excellent anti-leukaemic efficacy associated with a reasonably low TRM rate of 17–26% at 1 year.

HSCT in elderly patients with AML

Promising results have also been reported in 368 patients older than 50 years with standard or high risk AML who received HSCT between 1995 and 2005 from a matched sibling or an unrelated donor [22]. Not the donor type, but the patients’ disease status at HSCT and the cytogenetic risk proved to be the major prognostic factors for overall and event-free survival. The cumulative incidence of relapse at 2 years was significantly lower in patients with standard vs. high risk cytogenetics (29% vs. 42%) and of patients transplanted in first CR vs. advanced AML (26% vs. 38%).

HSCT for acute lymphoblastic leukaemia

In ALL, the indication for HSCT in first remission is widely accepted for patients with one or more unfavourable prognostic factors associated with a survival probability of less than 40% with chemotherapy alone. Criteria for high risk ALL include (1) detection of a Philadelphia chromosome or t(4;11); and/or (2) a particular immunophenotypically defined subtype (e.g. pro-B-ALL, early or mature T-ALL); and/or (3) a high white blood count at presentation; and/or (4) delayed response to chemotherapy [23].

ALL in first complete remission

A recent meta-analysis of seven published studies confirmed a significant survival benefit for HSCT in patients with high risk ALL with a matched sibling donor compared with no donor [24]. Importantly, outcome after HSCT from a matched related or unrelated donor is considered equivalent leading to overall survival rates of 40–45% at 5 years in retrospectively analysed series of patients with high risk ALL [25–27]. Use of a mismatched donor, however, resulted in poor survival because of excessive TRM. As time from CR to HSCT also had a negative impact on outcome, patients without a well-matched donor may benefit from a protocol including alemtuzumab for in vivo T-cell depletion [28] or use of umbilical cord blood (UCB) (see below).

The role of HSCT in standard risk ALL is still controversial. The MRC/ECOG study as the only large, genetically randomized trial demonstrating a survival advantage for standard risk ALL in patients is difficult to interpret, because all patients above the age of 35  were allocated to the high risk group. Because of the long accrual period of 13 years, younger patients of the no donor group did not participate in recent therapeutic progress by intensification of conventional chemotherapy [29]. As late morbidity and mortality are more pronounced, and the quality of life seems to be poorer in HSCT than chemotherapy-treated patients, young patients with standard risk ALL should not receive HSCT in first CR, as long as they experience an adequate decline of minimal residual disease (MRD) within the first months of chemotherapy. In contrast, a more than 90% relapse rate at 3 years for standard risk patients with a persistently high MRD niveau at ≥10−4 has been demonstrated by the GMALL study group and others [30]. Thus, most of the actively recruiting protocols have adopted MRD monitoring to further stratify the heterogeneous group of standard risk patients.

Relapsed ALL

HSCT offers long-term survival rates of 20–30% to patients with ALL in second CR. As treatment decisions are increasingly based on MRD status, a molecular relapse defined as reappearance of MRD above 10−4–10−3 should also trigger HSCT [23].

Selection of the stem cell source

Besides bone marrow (BM), G-CSF mobilized peripheral blood stem cells (PBSC) and UCB are increasingly being used [8]. Each stem cell source has its specific advantages and limitations.

Peripheral blood stem cells vs. bone marrow

In the setting of matched sibling HSCT, the use of PBSC and BM has been evaluated in numerous randomized trials including more than 1100 patients with hematologic diseases. In conclusion, PBSC lead to rapid engraftment and was associated with a reduced relapse rate, but a significantly increased risk of overall and extensive chronic GVHD compared to BM (68% vs. 52% and 47% vs. 31% at 3 years, respectively). Overall and disease-free survival were statistically significantly improved only in patients with late-stage disease. Thus, PBSC is the preferred stem cell source for aggressive and advanced hematologic malignancies.

Chronic GVHD has been identified as one of the most important prognostic factors for long-term survival of patients with high risk diseases because it is inextricably linked to a strong graft-versus-leukaemia effect. Extensive GVHD, however, is the most serious long-term complication of HSCT leading to substantial non-relapse mortality [31]. For patients undergoing HSCT for non-malignant conditions such as severe aplastic anaemia or hemoglobinopathies, the graft-versus-leukaemia effect is not applicable and BM should be chosen.

Importantly, the transplanting physician or donor registry must also take into account the preferences and potential risks for the stem cell donor associated with general anaesthesia required for BM harvest or side-effects of G-CSF-induced PBSC mobilization and apheresis.

Umbilical cord blood stem cells

UCB is obtained without any risk and rapidly available once a unit has been identified as suitable for a particular patient. The first UCBT was performed from a HLA-matched sibling donor in a child with Fanconi anaemia by Gluckman et al. [32]. The feasibility of unrelated UCBT has been proven in two larger series [33,34]. By now, UCBT has emerged as an effective and safe alternative stem cell source for children and adults who lack a well-matched adult donor.

UCB unit selection is based on HLA matching and cell dose. Unlike in unrelated adult donor transplants, no benefit of high-resolution typing could be demonstrated for UCBT in retrospective studies. HLA disparity should not exceed two of six parameters defined by HLA-A, and -B antigen and HLA-DRB1 allele typing to avoid graft rejection. The low incidence of GVHD with UCBT despite HLA mismatch may be attributable in part to the immaturity and lower number of immunocompetent T cells [35].

To date, randomized controlled trials comparing unrelated UCBT to BMT have not been performed, but large retrospective reports have compared outcomes for adults with acute leukaemia. The mostly myeloablative TBI-based conditioning regimens reliably lead to engraftment rates of >90% as long as cell dose and HLA matching thresholds are respected. In the report from the US Center for International Blood and Marrow Transplant Research [36] and from Eurocord [37], there was a significantly higher incidence of treatment-related deaths with the use of UCB compared to matched unrelated BM, likely related to slow engraftment and increased early infectious complications, which did not translate into a relevant difference in overall survival.

Recently, a retrospective analysis of the Japan Marrow Donor Program and the Japan Cord Blood Bank Network yielded more detailed information on disease-specific outcomes [38]. They included 999 AML and 691 ALL patients undergoing UCBT or BMT from unrelated donors performed between 2000 and 2005. For patients with AML, probability of overall survival (48% vs. 59%) and leukaemia-free survival (42% vs. 54%) at 2 years was significantly lower in the 173 UCB vs. the 311 BM recipients. In contrast, no difference was found for patients with ALL (52% vs. 53% and 46 vs. 44%, respectively). It is important to note that the UCB recipients with AML were likely to have more advanced disease associated with a higher relapse rate than the BM recipients. In view of these data, the donor search process for BM or PBSC and UCB from unrelated donors should be started simultaneously in patients with acute leukaemias where the time factor is critical.

As the most important factor influencing outcome is the cell dose, a minimum of 3 × 107 total nuclear cells/kg at collection is recommended. Thus, cell dose remains the most significant obstacle for adults – both in terms of limiting access to UCB and in terms of success after UCBT. Strategies to improve engraftment include intrabone injection of cord blood cells, co-infusion of mesenchymal stem cells and RIC [17]. In addition, the use of double UCB units has been explored with promising results [39,40]. In a recent phase I trial, ex vivo expansion of one of two UCB units has been shown to accelerate neutrophil recovery [41].

Novel approaches in HSCT

Recurrence of the primary disease is still the leading cause of death in patients with acute leukaemia, regardless of the stem cell source used. Thus, pharmacological or immunological approaches are being explored as pre-emptive therapy or in molecular relapse to prevent hematologic relapse after HSCT. Promising results have been achieved with targeted therapies such as imatinib in Philadelphia chromosome positive ALL [42], epigenetic modifiers such as 5-azacitidine in AML [43] or donor lymphocyte infusions in both AML and ALL [18]. As native CD3+ DLI may potentially induce lethal GVHD, phase I/II trials using CD3/CD28 activated T cells, natural killer cells and CD8-depleted T cells are currently underway [44].


No conflicts declared.