Indications, Outcomes, and Donor Selection for Allogeneic Hematopoietic Cell Transplantation for Hematologic Malignancies in Adults

Background

Early reports of leukemia-free survival, following human leukocyte antigen (HLA)-identical sibling hematopoietic cell transplantation (HCT) for refractory acute leukemia, was reported in the 1970s. Since then, allogeneic HCT has evolved into an effective and increasingly safe, curative, and intent therapeutic modality for hematologic malignancies. The indications and applicability of HCT for specific diseases change over time, especially as non-transplant treatment options and our understanding of disease biology improve. For example, in chronic myelogenous leukemia (CML), molecularly targeted therapy has almost completely replaced HCT since the early 2000s. At the same time, developments in HCT, such as the advent of reduced intensity conditioning (RIC), the use of mismatched, haploidentical-related donor and cord blood grafts, and improvements in supportive care, have increased the scope and applicability of allogeneic HCT. Allogeneic HCT is increasingly used in older patients, ethnic minorities, those for whom fully HLA-matched donors are not available, and those in whom fully HLA-matched donors are used as a salvage strategy after relapse. Transplant-related mortality (also referred to as non-relapse mortality [NRM]), despite declining over the years, remains an important challenge; posttransplant relapse has emerged as a major therapeutic focus. General principles, indications, donor selection, and optimal timing of allogeneic HCT for hematologic malignancies in adults and post-HCT outcomes are discussed in this chapter.

Allogeneic HCT involves administration of a conditioning (or preparative) regimen of chemotherapy (with or without radiation) and immunosuppressive medications, followed by infusion of donor hematopoietic progenitor cells. Most patients then receive several months of immune suppression therapy to prevent or treat graft-versus-host disease (GVHD). The purpose of the conditioning regimen is twofold: to eradicate malignant cells and to eliminate host immune cells (capable of rejecting even HLA-identical sibling donor cells). The ability to restore hematopoiesis with donor hematopoietic progenitor cells permits the administration of substantially higher (myeloablative) doses of cytotoxic therapy, than is otherwise possible. Additionally graft-versus-malignancy (GVM) effects, conferred by alloreactive donor immune cells, contribute substantially to cancer eradication and relapse prevention.

Patient Population

The Center for International Blood and Marrow Transplant Research (CIBMTR) collects data on HCT outcomes from over 500 worldwide participating centers and on all allogeneic transplants in the United States. Approximately 83,700 allogeneic HCT were performed in the United States between 2010 and 2018. In 2018, approximately 2371 matched sibling donors, 4309 unrelated adult donors, 1503 haploidentical donors (defined as a related donor with two or more antigen mismatches), and 466 cord blood transplants were performed in the United States, including approximately 7500 in adult patients. Hematologic malignancies are the most common indications for allogeneic HCT. Acute myeloid leukemia (AML), myelodysplastic syndrome/myeloproliferative neoplasms (MDS/MPN), and acute lymphoblastic leukemia (ALL) currently constitute the three most common indications of allogeneic HCT ( Fig. 105.1 ).

In recent years, improved supportive care and the use of RIC have led to an increase in allogeneic HCT for older adults. Only 4% of allogeneic HCT-recipients, from 1987 to 1992, were older than 50 years. In 2018, 45% were older than 55 years and 20% were 65 years or older. The use of allogeneic HCT in patients 70 years and older in the United States increased substantially in the past decade, from 5 (0.1%) in 2000, to 283 (3.85%) in 2013. Allogeneic transplantation, in patients without HLA-identical siblings, was facilitated by establishment of large unrelated donor registries around the globe. From 1987 to 1992, less than 10% of HCTs, for hematologic malignancies, used unrelated donors; in 2018, this figure was approximately 80%. Increased utilization of HCT among older adults, increasing use of unrelated donors, and more recently, haploidentical-related donors have been the main reasons for the steady growth in allogeneic HCT over the last decade. In 2011, approximately 300 haploidentical-related donor transplants were reported to the CIBMTR, compared with nearly 1500 transplants in 2018. The number of cord blood transplants declined concomitantly from a high of 816 in 2011, to approximately 500 in 2018. Over the years in the United States, the increase in the number of allogeneic HCT has been most striking in African Americans, while the number of allogeneic HCT for all indications increased by 7% from 2010 to 2018 for White people, the growth was 63% for African Americans, and 21% for Hispanics. A threefold increase occurred between 2010 and 2018 in haploidentical HCT; in 2018, approximately 20% of haploidentical HCT were performed in African Americans.

Conditioning Regimens

Historically, conditioning regimens included myeloablative doses of cytotoxic drugs—with or without radiation—intended to provide disease control and induce host immunosuppression (to prevent graft rejection). Myeloablative conditioning (MAC) regimens for hematologic malignancies often involve a combination of cyclophosphamide (commonly 60 mg/kg/day for 2 days) and either busulfan (typically 3.2 mg/kg/day intravenously for 4 days or a similar pharmacokinetic guided dose) or, historically, conditioning regimens included total body irradiation (TBI) (5 to 15 Gy, single or fractionated doses). Posttransplant survival rates are broadly similar with either of the above approaches, although an advantage for pharmacokinetic-guided busulfan regimens has been shown for myeloid malignancies. In a variety of regimens, other drugs such as fludarabine, etoposide, melphalan, and thiotepa are sometimes added to or substituted for cyclophosphamide and/or busulfan in efforts to provide better, generally disease-specific, antineoplastic activity. An Italian phase III randomized study in AML patients compared the two common MAC regimens (intravenous busulfan plus cyclophosphamide [BU CY], with busulfan plus fludarabine [Flu BU]) and demonstrated lower NRM (with busulfan plus fludarabine), without affecting relapse risk. In a study randomizing AML patients to high-dose cyclophosphamide and either 12 Gy or 15.75 Gy TBI as conditioning, a lower relapse risk in the 15.75 Gy TBI group was offset by higher NRM risk leading to similar OS. A more recent CIBMTR study also reported that higher doses of TBI over the standard dose (12 Gy) were associated with worse NRM and offered no survival benefit.

Toxicity associated with MAC regimens can be minimized and efficacy can be improved with careful pharmacokinetic monitoring of certain drugs, for example, busulfan. Another strategy for lowering NRM is reducing the dose-intensity of the conditioning regimen. This approach uses lower doses of cytotoxic drugs and/or radiation to facilitate donor cell engraftment and relies more on GVM effects to eradicate malignant cells. The lower doses of cytotoxic agents produce less host tissue damage and inflammatory cytokine secretion, resulting in lower rates of regimen-related morbidity and mortality. In patients ineligible for traditional MAC regimens, the use of RIC regimens has greatly increased the applicability of allogeneic HCT because of age or comorbidities. The development of RIC regimens has led to a substantial increase in the utilization of HCT in older patients ( Fig. 105.2 ).

Consensus criteria developed by the CIBMTR and the European Society for Blood and Marrow Transplantation (EBMT) distinguish between myeloablative, reduced intensity and nonmyeloablative regimens based on the likelihood of the regimen to produce toxicity in the recipient marrow ( Table 105.1 ). MAC regimens produce profound pancytopenia and are usually fatal in the absence of stem cell rescue. Nonmyeloablative regimens cause brief and often less severe pancytopenia, and in the absence of donor stem cell rescue, autologous hematopoietic recovery is likely to occur. RIC regimens are an intermediate category that does not fit well in either of the above, since they produce pancytopenia that may recover without stem cell rescue. RIC regimens induce less immune compromise in the immediate post-HCT setting as the duration and depth of neutropenia are reduced and host-derived immunocompetent cells are not immediately eliminated.

It is generally accepted that RIC regimens allow HCT in some patients who are ineligible or not offered HCT with MAC. For example, in the United States, the proportion of transplants performed in patients 60 years and older increased from less than 5% before 2000, to ~34% in 2018. Two-thirds of the latter patients received RIC regimens. However, RIC regimens are increasingly used in patients of all ages who receive both related and unrelated donor transplants, as is shown in Fig. 105.2 . Despite the growing popularity of RIC regimens, large retrospective database studies from the EBMT and CIBMTR (summarized in Table 105.2 ) have generally failed to show a definite survival advantage resulting from the use of such conditioning regimens. Although RIC regimens allow a potent graft-versus-leukemia response to occur with lower toxicity, there are concerns about a higher risk of disease relapse following RIC regimens. Since the distribution of conditioning regimens in registry studies reflects physician choice or patient selection bias, retrospective data are of limited value and randomized trials have addressed this question. The prospective US Bone Marrow Transplant Clinical Trial Network (BMT CTN) 0901 trial randomized patients with myeloid malignancies eligible for myeloablative conditioning to RIC versus MAC regimens. In this trial, a superior relapse-free survival was evident for MAC after enrolling 272 out of the planned 356 patients, leading to early closure of the study. Notably, an overall survival (OS) advantage did not emerge. A prospective randomized European trial (RICMAC) showed similar 2-year, relapse-free, and overall survival for busulfan-based RIC versus MAC in patients with MDS and secondary AML.

Donor Factors

NRM rates are generally lower after HLA-identical sibling than after unrelated donor or alternative donor transplants because of less graft failure, faster immune reconstitution, and less GVHD. However, considerable progress has been made in reducing NRM rates after both HLA-identical sibling and alternative donor HCT. For matched sibling donor HCT (for AML in CR1), the incidence of NRM at 1 year decreased from 23% from 1985 to 1989, to about 11% from 2000 to 2004, and down to 6.3% in 2018. The 1-year incidence of NRM for unrelated donor HCT (for AML in CR1) from 1990 to 1994 was 34%, which decreased to 26% from 2000 to 2004, and down to 9.5% in 2018. Among alternative donor transplants, those from more closely HLA-matched donors tend to have lower risks of GVHD and NRM. Donor-recipient HLA mismatching is associated with increased risk of posttransplant complications, including graft rejection, acute and chronic GVHD, and mortality; risks increase progressively with multiple HLA mismatches. With modern molecular HLA typing techniques (allowing selection of more closely HLA-matched donors) and current GVHD prevention strategies, the difference in outcomes between HLA-matched sibling and unrelated donor transplantation has narrowed ( Fig. 105.3 ). Adoption of molecularly defined, HLA matching techniques, calcineurin inhibitor–based GVHD prophylaxis, fungal prophylaxis with azoles, leukocyte reduction of blood products, newer assays for viral reactivation, pharmacokinetic testing of conditioning agents, and posttransplant cyclophosphamide for induction of allogeneic tolerance are some of the major innovations that have impacted NRM.

Graft Sources

The sources of hematopoietic cells for transplantation, historically bone marrow from donors, now also include peripheral blood hematopoietic cells, collected by leukapheresis and umbilical cord blood. The composition of grafts from bone marrow, peripheral blood, and cord blood varies in terms of the proportion of pluripotent stem cells, the proportion of lineage-committed late progenitor cells, and in the proportion and characteristics of immune-reactive cells. Bone marrow is the primary source of allogeneic donor cells for transplantation in children, and peripheral blood is the primary source in adults ( Fig. 105.4 ).

Several randomized trials have compared matched-sibling bone marrow versus peripheral blood grafts in patients with hematologic malignancy, following MAC regimens. In general, these studies showed no difference in the rates of acute GVHD between the two graft sources and showed no difference between higher risk of chronic GVHD with peripheral blood grafts. While survival outcomes were similar in most studies, the US and Canadian randomized studies reported a disease-free and overall survival advantage, respectively, favoring peripheral blood grafts. A meta-analysis, using individual patient-level data from nine randomized trials of related donor peripheral blood versus bone marrow transplants suggested that patients receiving peripheral blood grafts had faster hematopoietic recovery, a lower relapse rate if transplanted for hematologic malignancy, and higher overall and disease-free survival if transplanted for advanced-stage disease. However, peripheral blood was also associated with a significantly increased risk of extensive chronic GVHD. Among unrelated donor transplant recipients, the prospective randomized BMT-CTN 0201 trial showed a higher risk of chronic GVHD in patients receiving peripheral blood grafts (53% versus 41% for marrow grafts). Notably, survival and the incidence of relapse and acute GVHD were similar, while graft failure was more common among recipients of marrow grafts. The relative importance of graft sources in patients receiving lower-intensity conditioning is not known, but a CIBMTR study comparing unrelated bone marrow versus peripheral blood grafts, following RIC, showed no difference between the two sources in terms of rates of acute or chronic GVHD, relapse risk, and survival outcomes.

Impact of Measurable Residual Disease

At the time of transplant, disease remission status is a strong predictor of survival post transplantation for hematologic malignancies. Evaluation of measurable or minimal residual disease (MRD) status, using sensitive techniques (multiparameter flow cytometry, sequencing), is one of the most important established prognostic tools in ALL and is now emerging to be an important predictive and prognostic marker in other hematologic malignancies, such as AML, myeloma, and B-cell lymphomas. In ALL, MRD is present in up to 30% to 50% patients in a traditional CR. Consideration of MRD impacts several clinical decisions in HCT. Among these are: assessing the suitability of proceeding to HCT versus additional consolidative therapy in patients who achieve a traditional CR but with residual MRD, the intensity of conditioning regimen used for HCT (RIC versus MAC), the role of pre-emptive posttransplant therapy to address relapse risk, and the need for HCT in those with an MRD-negative remission.