Transplants should preferentially be performed on time and not reserved for relapsed disease or complete remission 2 (CR2) in high-risk patients. Although several reports of successful HCT in CR2 with a curative potential of 25% to 30% have been reported, such data are highly selective and should not influence a decision against earlier HCT, especially at experienced transplant centers. Moreover, the risk of relapse after URD HCT is higher for patients with unfavorable cytogenetics, irrespective of whether the transplant is performed in CR1 or CR2. Since the probability of achieving CR2 after chemotherapy in this group of patients is low, early HCT provides the optimal approach.⁹, ¹⁰, ¹², ¹³

Several clinical features at presentation are associated with particularly poor outcome for AML patients. Age over 60 years, hyperleukocytosis, secondary AML (after antecedent myelodysplasia), or treatment-related myeloid neoplasm and, in some series, two induction cycles rather than one to attain CR1 are clinical factors that portend a worse outcome. As a result, many investigators proceed to offer HCT to such individuals.⁹, ¹², ¹³

Cytogenetics, however, remains the most robust prognostic marker for risk stratification of AML at the time of diagnosis as well as in selection of post-remission treatments. On the basis of specific structural and numerical cytogenetic abnormalities, AML patients are divided into favorable, intermediate, and adverse risk groups (Chapter 75). A notable exception is the favorable cytogenetic risk group, in which HCT did not further improve outcome.

In recent years, a variety of novel molecular markers including mutations in the FLT3 (FMS-like tyrosine kinase) gene, the NPM1 (nucleophosmin) gene, and the CCAAT/enhancer binding protein alpha (CEBPA) gene have refined the risk stratification of intermediate-risk AML. Several investigators have used combinations of molecular markers to improve the patient’s risk assessment in cytogenetically normal AML.⁸

The most challenging group of patients with de novo AML falls within the intermediate-risk category.⁸, ¹² Intermediate-risk patients who achieve CR1 after chemotherapy have a 50% probability of disease recurrence without HCT, and the probability that CR2 can be attained is low. With advances in the molecular classification of AML such as NPM1, FLT3-internal tandem duplication (ITD), CEBPA, and KIT, the indications for HCT can now be extended to approximately 40% of patients with AML with no chromosomal abnormalities.⁹, ¹⁰, ¹¹, ¹², ¹³ In a recent meta-analysis by Koreth et al.¹⁴ for the transplant period 1982 to 2006, a significant overall survival benefit was reported for patients with poor-risk and intermediate-risk AML who received HCT in CR1. NPM1 positivity has been considered a good-risk feature. More recently, it has been shown that only the subgroup with wild type isocitrate dehydrogenase 1 and 2 (IDH1/IDH2) along with NPM1 mutations are expected to do well.¹⁵ Thus, the prognostic impact
of known well validated mutations continues to evolve as new mutations are identified and the interactions of these mutations are understood.

Importantly, the outcome of transplantation appears to be comparable for recipients of URD vs. MRD transplant. With improvements in transplant procedures including better donor selection (with high-resolution HLA- typing), excellent supportive care, and pre-emptive therapy in high-risk patients, we anticipate that outcomes will continue to improve.³, ¹⁶, ¹⁷, ¹⁸, ¹⁹, ²⁰, ²¹

The results of 4 comparative studies together with meta-analyses of cord blood transplant (CBT) vs. URD transplant, the results of CBT appear to be as promising as those of matched URD transplant in adults with hematologic malignances.¹⁰, ¹², ²⁰, ²² In the absence of an HLA-matched donor, both CBT and haploidentical-stem cell transplantation (SCT) strategies (center dependent) are suitable options to treat high-risk patients. An upcoming blood and marrow transplant (BMT)-clinical trials network (CTN) study (BMT-CTN 1101; clinicaltrials.gov NCT01597778) will compare haploidentical-SCT with CBT.

Relapse is the major cause of treatment failure after HCT in patients with AML¹⁰, ²³; new approaches to prevent disease recurrence are being explored. A question of practical importance is whether patients undergoing transplantation in CR1 will benefit from post-remission chemotherapy prior to transplantation. There are no prospective studies demonstrating that any form of post-remission chemotherapy further reduces the risk of post-transplant relapse. However, results from two retrospective registry analyses suggest no benefit from adding further consolidation chemotherapy prior to HCT; in both studies (transplant period 1980 to mid-1990s), most patients were young and received full-intensity conditioning MRD HCT.⁹, ¹², ¹⁴,²⁴ The role of post-remission consolidation chemotherapy prior to RIC HCT in HR AML has not been adequately addressed.¹³, ²⁵ In a recently published CIBMTR study, relapse occurred in 37% to 40%, and was higher in patients receiving RIC regimens. These findings highlight the importance of the depth of remission pre-SCT, especially in recipients in the RIC group.¹⁸ The efficacy of current chemotherapeutic regimens to achieve minimal residual disease negativity needs to be balanced against the risk of increasing co-morbidities prior to HCT.

Given the expanding use of RIC conditioning and the argument that treatment intensity may be important for cytoreduction prior to SCT, assessing the impact of pre-transplant consolidation therapy in AML CR1 on outcomes post RIC HCT is an important clinical question that remains unanswered. There is an ongoing CIBMTR analysis to determine if consolidation chemotherapy prior to RIC regimen for AML CR1 improves survival and relapse risk after RIC HCT.

A recent summary of several clinical trials comparing HCT with alternative therapies for patients with AML CR1 showed benefits of HCT over alternative therapies.⁸, ⁹, ¹³, ²⁶ Studies showed lower relapse rates for allograft recipients compared with those who received auto-SCT or further chemotherapy. None of the studies demonstrated an advantage for any of the alternative modalities (auto-SCT or further chemotherapy) as compared with HCT.

Conditioning regimens prior to HCT serve two purposes. One is the suppression of host immunity to allow for donor cell engraftment, and the other is the ablation of underlying malignancy. Ideally both these tasks need to be accomplished with minimal extramedullary toxicity. Conditioning regimens can be broadly classified as myeloablative (MST), RIC, and non myeloablative (NMA), based on the intensity of conditioning. MST regimens cause irreversible cytopenia and hence stem cell support is necessary. RIC regimens cause profound cytopenia and should be given with stem cells but cytopenia may not be irreversible. NMA regimens can theoretically be given without stem cell support.

MST HCTs are associated with decreased relapse but with increased transplant-related mortality (TRM). We now know that myeloablation is not as critical for engraftment as is immunosuppression. Reduced-intensity immunosuppressive approaches such as antithymocyte globulin (ATG), alemtuzumab, rituximab, and others can be used for transplant conditioning. There is a balance between reducing TRM by reducing intensity, and risking more relapse as the regimen is weakened.⁶, ²⁷

No study to date has shown superiority of RIC to MST in AML⁴, ²⁷, ²⁸, ²⁹, ³⁰; we recommend reserving RIC (but not truly NMA) for patients thought to be ineligible for the more intense MST regimen. The importance of disease control at the time of transplant, especially for those patients undergoing the RIC condition, appears to have paramount importance when evaluating outcomes. In Martino’s large-scale retrospective analysis of AML patients undergoing SCT, those patients transplanted in CR1 had similar rates of relapse (MST: 27% vs. RIC: 32%).³¹ A prospective study randomizing patients to MST or RIC HCT regimens based solely on age of < or >50 years for AML patients in CR1 revealed similar non relapse mortality of 19% for MST and 20% for RIC with equivalent relapse and survival.³² These data suggest that MST conditioning may not be needed for those AML patients in CR1 at the time of HCT. Ongoing prospective randomized clinical trials (BMT-CTN 0901; clinicaltrials.gov NCT01339910) will answer this question more definitively.

More recently, the investigators from City of Hope National Medical Center (COHNMC) reported their extended experience with 71 patients with refractory AML who were treated with HCT during a period of 22 years with 82% receiving grafts from MRD. The median age of these patients was 37 years (range 2 to 62 years) and the median follow-up was 1 year. Three-year disease-free survival (DFS) was 29% with a relapse rate of 54%. Cytogenetic data were available on 51 of the patients treated at COHNMC. Three patients had favorable, 27 intermediate, and 21 unfavorable cytogenetics. Those patients with favorable or
intermediate cytogenetics had DFS of 44% vs. only 18% in those with unfavorable cytogenetic features.³³

A score to predict transplantation success was developed in a retrospective survey of 1,673 patients who underwent a MST HCT for AML in relapse. In patients with low score the 3-year overall survival (OS) was 42% and those with a very high score had a dismal prognosis with only 6% alive at 3 years. The high score refers to the same group of patients in whom attempts to reinduce second remission are likely to fail. Patients with a short CR1, poor cytogenetics, and/or circulating blasts present a major clinical challenge.³⁴, ³⁵

An interesting strategy of sequential cytoreductive chemotherapy, immediately followed by RIC and prophylactic donor lymphocyte infusion (DLI) in refractory and early relapse patients, was evaluated.³⁶ This German protocol known as the FLAMSA (fludarabine, Ara-C, amsacrine)-RIC approach was reported in a prospective multicenter study involving 103 patients. The median survival of 16.4 months and the 4-year estimated OS of 32% were promising.³⁶ This approach requires an HLA-matched donor available just when relapse is diagnosed. Therefore, for AML patients in CR1 even if HCT is not scheduled, a search for a donor should be initiated to enable sequential transplantation in case of early relapse.

Risk stratification is important to identify patients with AML who might benefit from HCT in CR1. An MD Anderson Cancer Center (MDACC) study analyzed 150 patients with AML and diagnostic cytogenetic abnormalities who underwent MST HCT while in CR1 to evaluate the prognostic impact of persistent cytogenetic abnormalities at HCT.³⁷ Three risk groups were identified. Patients with favorable/intermediate cytogenetics at diagnosis (n = 49) and patients with unfavorable cytogenetics at diagnosis but without a persistent abnormal clone at HCT (n = 83) had a similar 3-year leukemia-free survival of 58% to 60% despite the higher 3-year relapse incidence in the latter group (32.3%, versus 16.8% in the former group). A third group of patients with unfavorable cytogenetics at diagnosis and a persistent abnormal clone at HCT (n = 15) had the worst prognosis, with a 3-year relapse rate of 57.5% and 3-year leukemia-free survival of only 29.2%. These data suggest that patients with AML and unfavorable cytogenetics at diagnosis and a persistent abnormal clone at HCT are at very high risk for relapse after HCT.³⁷,³⁸ These patients should be considered for clinical trials designed to optimize conditioning regimens (e.g., clofarabine-based regimen, see below) and/or to use preemptive strategies (e.g., chemotherapy [azacitidine], DLI, early withdrawal of immunosuppression if no GVHD, etc.) in the post-transplant setting, aimed at decreasing relapses.

Although the results of HCT in CR1 indicate an overall improved outcome compared with chemotherapy, many patients do not undergo transplantation while in CR1 based on either non-severe disease symptoms and/or on the patient desire, or mainly due to physician/patient preferences.

The currently available data suggest that patients with AML with untreated first relapse can expect an outcome following MST HCT similar to that of those patients proceeding to transplantation during CR2.¹⁰ Studies evaluating patients who receive an HCT in second or subsequent relapse, or third or later remissions following standard therapy, have demonstrated very few long-term survivors.

Although HCT can be curative, even in advanced disease, treatment failure is commonly manifested by relapse of disease, for which treatment is successful in only a minority of patients.²³ Detection of minimal residual disease in post-transplant surveillance is felt to be a necessary component of any strategy.³⁹, ⁴⁰ Past strategies for relapse prevention have focused on the use of DLI with variable success. Peritransplantation intervention and maintenance therapies (e.g., azacitidine) are under current investigation.⁴¹, ⁴²

Traditionally, the detection of impending relapse in the post-transplant period had been based on donor chimerism analysis (e.g., poor donor chimerism has been associated with increased risk of relapse), but the lack of specificity remains a problem. In AML, the development of post-transplant minimal residual disease strategies is more complex because of the genetic heterogeneity of these disorders. Nevertheless, for some genetic subtypes, for example, reciprocal translocations or NPM1-mutated AML, minimal residual disease monitoring in the post-transplant period has already been realized.³⁹, ⁴³, ⁴⁴, ⁴⁵

Walter et al.²⁴ investigated 99 AML patients receiving MST HCT CR1 by 10-color flow cytometry before transplantation. minimal residual disease-positive patients had lower 2-year estimates of survival (30.2% versus 76.6%) and higher 2-year estimates of relapse (64.9% versus 17.6%), compared with minimal residual disease-negative patients. After adjustment for other prognostically relevant parameters, a positive minimal residual disease status pre-HCT was significantly associated with increased overall mortality and relapse relative to minimal residual disease-negative HCT. Kebriaei et al.⁴⁶ found a trend toward improved outcomes in 68 patients in cytogenetic remission compared with those with residual cytogenetic abnormalities at the time of HCT.

In the conservative treatment setting of AML, many minimal residual disease studies confirmed the potential of real-time PCR to predict the relapse risk in patients with reciprocal rearrangements. Some studies investigated the potential of molecular minimal residual disease monitoring for patients with reciprocal rearrangements also in the post-transplant period.³⁹, ⁴³, ⁴⁷

Incorporating novel agents in the conditioning regimen with potent antileukemic activity is a strategy that is currently under investigation.⁴⁸ Clofarabine is a novel nucleoside analog that has potent antileukemic activity. It has shown efficacy in refractory disease as well as upfront scenarios. Geyer et al.⁴⁹ combined total-body irradiation (TBI) with clofarabine and cytarabine for HCT in pediatric patients with refractory or multiply relapsed leukemia in a phase I/II study. The early results show the probability of a 1-year progression-free survival of 57.4%. Andersson et al.⁵⁰ reported on a clofarabine/fludarabine/busulfan-based RIC regimen for advanced MDS and AML in 42 patients, which included children. In this high-risk patient population, 20 patients are alive at median survival of 23 months, which is an encouraging early result and warrants further studies.

Targeted immunotherapy post-SCT: One of the strategies for the prevention of relapse of AML/MDS post-SCT is to use targeted immunotherapy as part of the conditioning regimen such as the employment of vaccines, and DLI. Additionally, natural killer cells
could be transferred for the treatment of leukemia relapse both during and after SCT.²³, ⁵¹, ⁵²

The posttransplant therapeutic approach has been limited to TKI in patients with BCR-ABL1-positive leukemia. However, other strategies, including the use of interleukins, monoclonal antibodies, immunomodulatory agents (bortezomib), DNA methyltransferase inhibitors, and histone deacetylase inhibitors, are currently being explored in AML.³⁹ The goal of these approaches is to treat minimal residual disease while minimizing adverse side effects and avoiding the impediment of donor cell engraftment. The use of hypomethylating agents (e.g., azacitidine) in patients with AML/MDS has drawn increased interest because of their demonstrated efficacy and relatively acceptable toxicity profile.⁴¹, ⁴²,⁵³

As our ability to monitor minimal residual disease post-HCT improves, along with identification of actionable molecular mutations, it will allow for the development of clinical trials aimed at prevention of relapse in a targeted fashion. Therapeutic interventions will be needed not only to prevent relapse but allow for simultaneous favorable modulation of the GVHD-graft versus tumor (GVT) balance.

Relapse after HCT for AML carries a very poor prognosis. Second HCT, although curative, is not an appropriate treatment option for a large number of relapsing patients (only 2% to 20% patients receive a second HCT), and efforts to increase the number of patients who may benefit from a second HCT are ongoing.²³

The development of novel agents is expected to gather momentum as new actionable mutations are identified. A mutation in FLT3 is well known to be a poor prognostic marker in newly diagnosed and relapsed AML. Specific inhibitors of FLT3 such as sorafenib, lestaurtinib (CEP-701), and quizartinib (AC220) have been studied in relapsed or refractory AML.¹⁰, ³⁹, ⁵⁴ FLT3 inhibitors need to be investigated post-SCT in patients with FLT3 + AML to prevent relapse, a most common cause of transplant failure.⁵⁵

Hypomethylating agents are active in high-risk MDS, and to a lesser degree in highly proliferative AML. 5-Azacitidine was shown in one study to prolong survival in elderly AML patients with low proliferative disease (expected disease course after HCT), and hence has drawn attention as an agent that may control low proliferative relapses after HCT.⁵³

MDS originates in hematopoietic stem or precursor cells, and a logical approach is, therefore, to replace the clonal MDS cells by healthy donor cells via HCT procedure. Successful HCT requires that the infused cells from the healthy donor establish themselves (engraft), and that the malignant cells of the patient’s disease are eliminated by graft-versus-tumor (GVT) effects.

Over the past decade the approach to treating MDS has undergone major changes, at least in part related to the Food and Drug Administration (FDA) approval of several drugs for MDS and the development of new concepts for HCT. The drug therapy has improved the prognosis of patients with MDS, with 35% to 65% of patients achieving clinically relevant hematologic responses which may last from a few months (for high-risk MDS) to several years (predominantly for low-risk diseases). Concurrently, the emphasis with HCT has shifted from high-dose therapy, aimed at maximum tumor cell kill, to low- or reduced-intensity conditioning, relying heavily on donor cell-mediated immune effects to eradicate the disease (GVT effect). This GVT effect eliminates residual cells of the patient’s disease via recognition of minor histocompatibility antigens.

In an attempt to improve chances for post-SCT success, patients with a high marrow myeloblast count, (e.g., >5-10%), typically will receive pre-SCT therapy with a hypomethylating agent or with more classical induction-type chemotherapy.⁵⁶, ⁵⁸, ⁶⁸ The clinical impression is that pre-SCT therapy may select chemosensitive patients since it appears that patients who do respond to such therapy have a superior outcome, and patients who fail to respond have a poorer outcome than do untreated patients.⁷⁵

The CR status at the time of transplant may impact relapse and TRM rates favorably; however, CR in chemotherapytreated patients with MDS is infrequent. Warlick et al.⁷⁶ studied 84 patients undergoing HCT, and observed a 1-year cumulative incidence of relapse of 18% in patients in CR compared to 35% for those transplanted with active disease. In a retrospective analysis of 49 MDS/AML patients treated with T cell-depleted grafts, the 2-year cumulative incidences of TRM were 23%, 38%, and 40%, respectively, dependent upon response status prior to HCT (responsive disease, untreated, and refractory).⁷⁷ Oran et al. also documented the benefit of minimal tumor burden at HCT in patients with high-risk MDS/AML undergoing RIC HCT. CR at SCT was associated with day-100 and 2-year TRM rates of 0% and 20%, respectively. Estimates of 2-year OS were 66% for those in CR, 40% for patients with active disease without circulating blasts, and 23% for patients with circulating blasts.⁷⁸

Nakai et al.⁷⁹ compared the outcomes of 283 adult patients with MDS (n = 215) or secondary AML (n = 68) who received (n = 188) or did not receive (n = 95) induction chemotherapy prior to HLA-identical sibling donor HCT, and showed no difference in survival rates between the two groups, with 5-year OS rates of 54% and 57%, respectively. Similarly, Scott and collaborators in Seattle failed to demonstrate a clear survival advantage for a cohort of patients studied retrospectively that received induction chemotherapy pre-transplant, although the risk of relapse was somewhat reduced.⁸⁰

In the absence of controlled data, the decision to use pre-SCT induction therapy should be made on an individual basis. The treatment strategy should also take into account the fact that promising novel drugs are becoming available, although it is unclear if this approach will change long-term outcomes.

Field et al.⁸⁵ presented data in 54 consecutive patients with MDS or chronic myelomonocytic leukemia (CMML) who underwent HCT from HLA-matched donors. Thirty patients had received azacitidine for one to seven (median 4) cycles, and 24 patients had not received azacitidine. At 1 year after HCT with conditioning involving high-dose busulfan and fludarabine, 47% of azacitidine-treated patients were alive, compared with 60% of patients not given azacitidine, but relapses were 20% and 32%, respectively. The authors concluded that outcome was consistent with a trend toward a lower incidence of relapse in azacitidine-treated patients.

In a study by De Padua Silva et al.⁸⁶ 17 patients with MDS underwent HCT after having received decitabine. At 1 year, 11 patients were alive, whereas six had died: four from disease progression, one from GVHD, and one with septicemia. Lubbert et al.⁸⁷ presented data on 15 similarly treated patients who then underwent HCT following RIC regimen. At the time of reporting, 6 patients were alive in remission, whereas four had died from progressive disease and five from transplant-related complications while in remission.

FHCRC recently reviewed results in 68 patients who underwent HCT for MDS or AML transformed from MDS (tAML). Thirty-five patients had received cytoreduction with azacitidine prior to SCT followed by either an MST (40%) or RIC (60%) regimen, and 33 patients had undergone induction-type chemotherapy followed by SCT with MST conditioning. The 1-year overall survival was 57% in the azacitidine group, and 36% in the group given induction chemotherapy. Although the risk of post-SCT relapse and NRM was lower in the azacitidine group, none of the differences were statistically significant.⁸⁸

Post-SCT relapse has remained a major cause of failure, particularly in high-risk MDS. Earlier studies had shown that intensification of the transplant conditioning regimen reduced the relapse frequency, though the associated increase in NRM may prohibit such use. It may be advantageous to allow for recovery from the conditioning-related effects, and then administer additional agents. As most of the post-transplant relapses occur within the first 1 to 2 years, addition of hypomethylating agents may have immunomodulatory effects that could potentially increase the GVT response.⁶⁸

In a phase I dose-finding study, de Lima et al.⁵³ at the MDACC started azacitidine at 6 to 7 weeks after HCT and determined that a dose of 32 mg/m² given for 5 consecutive days was well tolerated. Forty-five patients with high-risk MDS/AML were treated and 1-year RFS was 58%. In addition, the authors noted that the incidence of chronic GVHD was lower in 5-azacitadine-treated patients than in historical controls. This observation was confirmed in a study at Washington University, which emphasized, in addition, that the beneficial effect on GVHD did not compromise the GVT effect.⁸⁹

Similarly, Platzbecker et al.⁴¹ reported a prospective study of patients receiving HCT for MDS and AML. If the donor chimerism fell below 80%, patients were offered treatment with 5-azacitidine. This event occurred in 25 of 59 screened patients at a median of 169 days, and azacitidine was initiated at 75 mg/m²/day for 7 days. Sixteen patients responded with an increase in donor chimerism or stabilization at 80%. Eleven patients were given additional 1 to 11 (median 4) cycles of azacitidine. Hematologic relapse eventually occurred in 13 patients (65%), but was delayed to days 56 to 558 (median 231). Therefore, preemptive azacitidine treatment may be effective in delaying hematologic relapse.