Biology of Disease Relapse in Myeloid Disease: Implication for Strategies to Prevent and Treat Disease Relapse After Stem-Cell Transplantation

INTRODUCTION

Allogeneic hematopoietic cell transplantation (HCT) is the only curative therapy for many patients with high-risk acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). While the risk of transplant-related mortality has decreased over the past 40 years, disease relapse remains one of the most significant adverse events following HCT, with up to 50% of patients with AML relapsing after HCT.¹ Outcomes for patients who relapse after HCT are poor.^1-3 The therapeutic benefit of HCT is believed to reside in part through an immune-mediated graft versus malignancy (GVM),⁴ and over the years, many attempts to improve outcomes have focused on investigating and exploiting this phenomenon. In this review, we will first review the biology of relapse after HCT, with a focus on recent discoveries. In the second part, we will evaluate strategies to prevent and treat relapsed disease post-HCT, highlighting novel molecularly targeted and immunologic treatments.

CONTEXT

• Key Objective

• Acute myeloid leukemia (AML) relapse after allogeneic stem-cell transplantation represents the most common form of treatment failure. This review summarizes recent findings about the mechanisms of AML relapse after transplantation, as well as therapies to prevent or treat post-transplant relapse.

• Knowledge Generated

• AML relapse after transplantation is associated with several factors, including active disease before transplantation, conditioning intensity, loss of antigen presentation by relapsing AML cells, and immune cell exhaustion. Studies of post-transplantation maintenance, either with hypomethylating agents or tyrosine kinase inhibitors, have shown some encouraging results. However, standard strategies to treat overt relapse with chemotherapy or donor lymphocyte infusion rarely result in long-term disease control.

• Relevance

• New therapies to prevent or treat AML relapse after transplantation are urgently needed. Strategies to reverse or circumvent immune escape by AML cells—for example, by using bispecific T cell–engaging molecules—may prove effective in this regard.

THE BIOLOGY OF RELAPSE AFTER ALLOGENEIC HCT

Leukemia Burden Pre-HCT

It has been long appreciated that the presence of active AML at the time of transplantation is associated with a higher risk of relapse. More recently, it has been reported that even low levels of minimal residual disease (MRD) at the time of transplantation lead to worse outcomes.⁵ Whether additional chemotherapy to eradicate MRD before HCT mitigates this risk has never been directly tested, but several lines of evidence suggest that aggressive chemotherapy before HCT may be of benefit. First, in CTN 0901, patients receiving myeloablative conditioning regimens had a lower relapse rate than patients who received reduced-intensity conditioning.⁶ Second, patients with high-risk or secondary AML who received induction and consolidation with (CPX-351) not only had better postinduction disease control than patients who received 7 + 3, but this translated to a lower post-HCT relapse rate as well.⁷ It is likely, therefore, that deeper remissions before HCT will lead to lower relapse risk as has been seen, for example, in multiple myeloma.

Given these findings, accurate MRD monitoring before and after HCT is crucial. Molecular MRD monitoring for mutated genes, fusion genes, and/or overexpressed genes can be performed via multicolor flow cytometry, quantitative reverse transcriptase polymerase chain reaction (PCR), digital droplet PCR, or next-generation sequencing (NGS).⁸ Given its improved sensitivity, reduced cost and turnaround time, ability to track multiple molecular markers and clonal hierarchy at once and guide molecularly directed therapies (ie, FLT3 inhibitor or isocitrate dehydrogenase [IDH] inhibitors), and utility in predicting clinical outcome in HCT, error-corrected NGS is becoming feasible for routine clinical MRD estimations before and after HCT.⁹

Disease Characteristics

Several cytogenetic abnormalities and gene mutations are predictive of relapse in patients with AML and MDS undergoing HCT. Poor-risk cytogenetics associated with a significantly increased incidence of AML relapse after HCT includes patients with monosomal or complex karyotypes as well as inv(3)(q21q26)/t(3;3)(q21;q26), del(5q), t(10;11)(p11-14;q13-23), t(6;11)(q27;q23), and abnormalities in chromosome 17p.^10,11 In MDS, abnormalities of chromosome 3 [ie, inv(3), t(3q)] or chromosome 7 [-7, del(7q)], either alone or as part of a complex karyotype, result in inferior leukemia-free survival after HCT.¹² At the gene level, AML patients with TP53 and FLT3-ITD mutations exhibit inferior outcomes with a higher risk of relapse.^10,13-15 Likewise, mutations in TP53 as well as TET2, ASXL1, RUNX1, and RAS are independently associated with an increased risk of relapse after HCT for MDS.^16-19

The biological mechanisms mediating the increased risk of relapse after HCT in patients with AML and MDS with the cytogenetic and molecular disease states discussed above remain poorly characterized. In addition to its role as a tumor suppressor, TP53 regulates several innate and adaptive immune responses including antigen processing and presentation, cytokine production, type 1 interferon signaling, and expression of immune inhibitory receptors.²⁰ Recently, the examination of bone marrow samples obtained from patients with MDS or AML with TP53 mutations demonstrated higher CD8+ T cell, ICOS^high/PD-1^neg regulatory T cell (Treg), and natural killer (NK) cell infiltration, with increased expression of immune checkpoints (PD-L1, TIGIT, and LAG3) and interferon-gamma (IFNg) signaling compared with other AML subtypes.^21-24

Downregulation of Tumor Cell HLA Expression

First reported in 2009, loss of mismatched HLA loci in AML cells is associated with relapse in haploidentical HCT (haplo-HCT) through impairment of donor T cell recognition.²⁵ This loss, observed in roughly one third of relapses after haplo-HCT, is copy number neutral, representing a form of acquired somatic uniparental disomy (ie, loss of a chromosomal region followed by replacement with its homologous copy). Overall, the expression level of HLA molecules is unchanged, including major histocompatibility complex (MHC) class I, which may limit the activation of antitumor NK cells.²⁶ Loss of mismatched HLA is observed in about one third of haploidentical transplants and more frequently occurs in patients with active disease at the time of transplant, increased donor T cell dose, and longer time to relapse, and in patients with chronic graft-versus-host disease (cGVHD), all factors that may lead to increased GVM and therefore selective pressure against mismatched HLA genes.^27-29 Genetic loss of mismatched HLA alleles has also been observed (albeit less frequently) in mismatched unrelated donor HCT,^30-32 matched unrelated donor HCT,^31-33 and mismatched related donor HCT.³⁴

In addition to loss of HLA genes, a wide variety of genetic changes occur in AML cells from patients who relapse after HCT. It has long been observed that chromosomal and other structural abnormalities disappear and emerge at relapse after HCT, reflecting evolution of the malignant clone.^32,35-39 More recently, panel sequencing of myeloid malignancy–associated genes has also demonstrated gain and loss of driver mutations in MDS and AML cells after HCT.^14,40-44 These changes resemble the spectrum of mutations seen after chemotherapy and, unlike the loss of HLA genes described above, are not specific for post-HCT relapse.^40,45-49 Somewhat surprisingly, given the role of alloimmunity in preventing relapse, mutations in genes that govern the immune response are not commonly seen in AML relapse after HCT.^31,40

At the level of gene expression, however, two groups have recently reported significant dysregulation of immune genes in AML cells at relapse after HCT.^40,50 These studies compared matched diagnosis and post-HCT relapse patient samples and found downregulated surface and RNA expression of MHC class II molecules and associated genes in 30%-50% of samples.^40,50 A higher dose of infused donor T cells correlated with a higher likelihood of HLA class II downregulation, and in both studies AML blasts with a low MHC class II expression failed to stimulate HLA-mismatched T cells in vitro, suggesting that downregulation of MHC class II genes contributes to the ability of AML cells to evade immune effectors in relapse after HCT. As observed in previous models of HCT, IFNg restored HLA class II expression, raising the possibility that IFNg treatment could resensitize AML cells to donor immune cell killing.^40,50,51 These results are consistent with those of smaller previous reports of MHC loss at relapse.^34,52

Inhibitory Immune Checkpoint Molecule Modulation

In addition to alterations in HLA expression on tumor cells, modulation of immune checkpoint molecules on both tumor cells and T cells has been described as another important mechanism of relapse post-HCT. Several groups have reported upregulation of immune-inhibitory genes such as PD-L1 (CD274), B7-H3 (CD276), and PVRL2 (CD155) on AML blasts relapsing after HCT compared with diagnosis.^50,53 Expression of immune-inhibitory genes on AML cells is associated with the upregulation of exhaustion markers on T cells after HCT, including PD-1, TIM-3, TIGIT, and KLRG-1.^50,53-57 Noviello et al⁵⁸ found increased evidence of T cell exhaustion in memory stem and central memory T cells from the bone marrow of patients who relapsed compared with controls, and these exhausted bone marrow–infiltrating T cells at relapse expressed a restricted T cell receptor (TCR) repertoire and impaired effector functions. The upregulation of inhibitory immune checkpoint markers has led to several therapeutic trials of immune checkpoint blockade to treat AML relapse after HCT, as described below.

PREVENTION AND TREATMENT OF RELAPSE POST-HCT

Established Strategies for Enhancing Donor Immunity Post-HCT

One of the simplest methods of enhancing donor immunity and treating relapse after HCT is early withdrawal of immunosuppression, which can restore donor hematopoiesis and transient remissions in some patients with frank relapse after HCT.^59-61 This strategy may be most effective in the setting of incomplete donor chimerism⁶¹ or as a strategy to prevent relapse after HCT for high-risk disease.^62,63 Conversely, the use of in vivo T cell depletion with antithymocyte globulin or alemtuzumab for graft-versus-host disease (GVHD) prophylaxis is associated with increased relapse risk in some studies, although not all.^64-66 Similarly, administration of fresh immune effector cells as a donor lymphocyte infusion (DLI) can be given to prevent relapse in high-risk patients (prophylactic or preventive DLI) or to treat overt hematologic relapse. In the prophylactic/preventative setting, DLI is associated with overall survival (OS) rates of 40%-70% with GVHD rates of 15%-60%, which compare favorably to expected outcomes in transplant of high-risk AML.^67-69 For patients with evidence of MRD after transplant, DLI can reduce the incidence of relapse, with OS rates of approximately 60% and grade II-IV acute GVHD (aGVHD) rates of 21%-28%.^70,71 In contrast, when used to treat overt relapse after HCT, DLI (usually administered with chemotherapy) is associated with worse outcomes, with OS rates at 2 years of approximately 20% and grade II-IV aGVHD rates of 22%-43% in large, registry-based retrospective analyses.^72-74 In these studies, response to DLI with or without chemotherapy was best in patients with late relapses, suggesting that patients with longer remissions may be more likely to benefit from graft versus leukemia (GVL). Conversely, patients with relapse < 6 months after HCT have particularly poor outcomes, with long-term OS of 5%-10%. Taken together, these observations suggest that DLI may be most effective when tumor burden is low and that in the setting of overt relapse DLI is of limited benefit.

Second HCT represents another option for some patients with MDS/AML relapse after an initial HCT. Registry studies suggest OS rates of 17%-49% and grade II-IV aGVHD rates of 26%-53% with this approach.^72,75,76 Use of myeloablative conditioning regimens is associated with high rates of nonrelapse mortality (NRM), ranging from 30% to 50%.^76-78 Although the use of a different donor may enhance the GVM effect by providing immune effector cells with more favorable alloreactivity, several retrospective studies have reported similar outcomes between a second HCT from the same and a different donor.^75,79,80 A 2018 European Society of Blood and Marrow Transplantation registry study comparing second HCT versus DLI for relapsed AML showed similar OS rates (15% for DLI v 19% for HCT). Although NRM (27% v 10%) and grade II-IV aGVHD rates (37% v 20%) were higher for the group receiving HCT, these differences were not seen when the analysis was restricted to patients receiving therapy in CR.⁷⁵ In summary, while a subset of patients have some benefit from DLI or second transplant, new treatment approaches are urgently needed for AML relapse after HCT.

Checkpoint Inhibitors

Given the efficacy of checkpoint inhibitors in solid tumor oncology and the evidence of T cell exhaustion in AML relapse after HCT, there has been considerable interest in the use of checkpoint blockade in the post-HCT setting. Early phase trials showed promising results with ipilimumab in patients with both lymphoid and myeloid malignancies relapsing after HCT.^81-83 In contrast, early trials of the PD-1 blocker nivolumab suggested a significantly increased risk of toxicities including severe GVHD in HCT patients.^84-87 Further study will be required to determine if the choice of checkpoint blockade agent or use of lower doses might reduce the toxicity seen so far with these drugs, but at this time concerns for toxicity have limited their widespread use after HCT.⁸⁸

Hypomethylating Agents

Based on preclinical studies suggesting their ability to mitigate GVHD and enhance the GVM effect, hypomethylating agents (HMAs) have been studied intensely in the post-HCT setting over the past 10-15 years.^89-92 Several small phase I/II studies have shown that HMAs may prevent relapse in patients with high-risk or MRD-positive MDS and AML post-HCT, with relapse rates ranging from 17% to 65%, relapse-free survival (RFS) rates of 46%-63%, and OS rates of 49%-77%.^90,93-97 Neutropenia and thrombocytopenia are common significant toxicities of HMA maintenance. While promising, these results await confirmation in larger, randomized trials. A number of trials have also evaluated the use of azacitidine in overtly relapsed disease.^99-103 Complete response (CR) rates in these studies range from 13% to 23% with a few long-term responses, although many of the patients received concurrent DLI. Thus, it appears that HMA treatment has a limited ability to enhance antitumor immune cell activity in patients with active disease after HCT and may be considered as maintenance therapy after HCT.

Histone Deacetylase Inhibitors

Histone deacetylase (HDAC) may promote GVM activity while reducing GVHD after HCT in the following ways: (1) increasing the expression of tumor antigens, MHC class I and II, and costimulatory molecules on malignant cells; (2) reducing the secretion of proinflammatory cytokines; (3) promoting the recovery of the intestinal barrier function after conditioning; and (4) expanding CD4+ Tregs.¹⁰⁴ HDAC inhibitor monotherapy and combination therapy have shown limited efficacy against R/R AML.^105-107 However, maintenance therapy with the pan-HDAC inhibitor panobinostat alone or in combination with a DLI after HCT for AML or MDS yielded an encouraging 2-year OS and RFS of 81% and 75%, respectively.¹⁰⁸ A randomized phase III trial to assess the efficacy of panobinostat maintenance therapy after HCT is ongoing (ClinicalTrials.gov identifier: NCT04326764).

Lenalidomide

Trials evaluating maintenance therapy with lenalidomide early (within 3 months) after T cell–replete HCT were unsuccessful due to the high rate of aGVHD.^109,110 More recently, combination therapy with azacitidine and lenalidomide in patients with AML or MDS who had relapsed after HCT yielded a CR rate of 40% (6/15; CR, n = 3; complete remission with incomplete hematologic recovery [CRi], n = 3), which was better than that in historical controls receiving only azacitidine.¹¹¹ Since only three patients developed aGVHD (grade II, n = 2; grade III, n = 1) upon receipt of the combination regimen, it is possible that azacitidine mitigates the lenalidomide-mediated exacerbation of GVHD.

Tyrosine Kinase Inhibitors

A number of studies have suggested that tyrosine kinase inhibitors have activity in preventing or treating relapse after HCT in AML patients with activating FLT3 mutations. Recently, results from randomized trials comparing sorafenib maintenance with placebo after HCT were reported. In the phase II Sormain trial, patients treated with sorafenib had significantly better RFS (85%) compared with patients who received placebo (53.3%).¹¹² Likewise, a significantly increased RFS was observed in the phase III trial (56.6% RFS for nonmaintenance v 78.9% RFS with sorafenib at 24 months post-HCT).¹¹³ Currently, the second-generation FLT3 inhibitor gilteritinib, which has shown single-agent activity in nontransplant relapsed refractory (R/R) AML, is also being studied as maintenance in FLT3-mutated AML.¹¹⁴ Finally, for patients with overt relapse after HCT, sorafenib monotherapy has shown modest activity in several smaller studies and case reports, with CR rates of 23%-38%, although these responses are generally short-lived.^115,116 Similar responses were reported for AML patients who relapsed after HCT and were treated with gilteritinib or quizartinib.^117,118

The first-generation FLT3 inhibitors such as sorafenib are relatively nonspecific for FLT3 and target other molecules and pathways such as c-Kit, platelet-derived growth factor receptor, vascular endothelial growth factor receptor, Janus kinase 2, and RAS/RAF/MEK pathway to induce direct killing of malignant cells.¹¹⁹ FLT3 inhibitors can also promote the GVM effect after HCT by inducing interleukin-15 (IL-15) production from FLT3-ITD+, but not non-ITD, AML cells via inhibition of the ATF4 transcription factor.¹²⁰ This IL-15 production promoted the GVM activity of donor CD8+ T cells in both mice and humans.¹²⁰ Furthermore, FLT3 inhibitors can modulate immunological responses by decreasing the number of immunosuppressive CD4+Foxp3+ Tregs and myeloid-derived suppressor cells (MDSCs) and inhibiting dendritic cell (DC) proliferation, maturation, and function.^121-123 Therefore, sorafenib may enhance GVM activity without increasing GVHD lethality because of its ability to reduce the numbers of immunosuppressive cell subsets (Tregs and MDSCs) and DCs while concomitantly enhancing the efficacy of leukemia-reactive CD8+ T cells via the release of IL-15.

IDH Inhibitors

Ivosidenib and enasidenib, small-molecule inhibitors of IDH 1 and 2, respectively, have been approved by the FDA for use as single agents in patients with R/R AML. Although limited information is available about the response rate in the 43 transplanted R/R AML patients treated with ivosidenib, 10 of 29 (34%) patients who relapsed with AML after HCT achieved a CR upon treatment with enasidenib.¹²⁴ Both ivosidenib and enasidenib are currently being tested as maintenance therapy for IDH1/2-mutant myeloid malignancies following HCT (ClinicalTrials.gov identifier: NCT03515512 and NCT03564821).

Venetoclax

Byrne et al¹²⁵ recently reported the outcomes for 21 patients with myeloid diseases who relapsed after HCT and were treated with venetoclax salvage chemotherapy in combination with a HMA (n = 16) or low-dose cytarabine (LDAC) (n = 5). An overall CR/CRi rate of 42.1% (n = 8/19) was reported for 19 evaluable patients with five responses in the HMA cohort and three in the LDAC group. Two studies testing venetoclax in combination with azacitidine as maintenance therapy after HCT in patients with AML are ongoing (ClinicalTrials.gov identifier: NCT04161885 and NCT04128501).

Tumor Vaccines

The period immediately after HCT may provide the best window for tumor vaccination because of the low tumor burden and the presence of fresh immune effector cells.¹²⁶ A pilot study evaluating the delivery of a Wilms' Tumor-1 (WT1) peptide-loaded donor-derived dendritic cell vaccine given concurrently with DLI to patients relapsing after HCT was safe and feasible with the evidence of WT-1–specific T cell responses.¹²⁷ More recently, Lichtenegger et al¹²⁸ reported antigen-specific responses and prolonged RFS compared with a historical control cohort (1,084 v 396 days) following vaccination of 10 patients with high-risk AML post-HCT with TLR7/8-matured DCs transfected with RNA encoding two AML-associated antigens, WT1 and PRAME, as well as CMVpp65. Although a few cases have been reported of overt AML relapse treated successfully with tumor vaccines,¹²⁹ it is likely that vaccination will prove more effective at preventing relapse than treating overtly relapsed AML.

Cellular Therapies Post-HCT

Many groups have attempted to improve outcomes after HCT by improving the alloimmune effect of donor cells toward residual recipient AML cells. One early approach was treatment with immune-stimulating cytokines such as interferon alpha, which has been reported to stimulate NK and T cell effector functions in preclinical studies.^130,131 While these treatments were well-tolerated, results were mixed and likely hampered by the short in vivo half-life of these agents in their unmodified form, and current efforts have focused on modified long-acting cytokine agonists such as ALT-803, described below.

A second approach to improving the GvL effect to prevent or treat relapse is through graft engineering, the transfer of specific immune cell subsets. Nikiforow et al¹³² reported a 43% CR rate (9/21; n = 11 with AML) with a 33% incidence of clinically significant GVHD upon infusion of donor lymphocytes depleted of CD25+ T regulatory cells. Since memory T cells cause significantly less GVHD than naive T cells, a phase I clinical trial of 15 patients evaluated the infusion of donor CD8+ memory T cells in patients relapsing after HCT (8 of 15 having AML).¹³³ Seven of the 15 patients achieved a CR, and only one patient developed grade II GVHD after the infusion.

In contrast to allogeneic T cells, NK cells do not cause GVHD and they have been shown to have anti-AML activity in the nontransplant setting where concurrent cytokine treatment is used to improve their duration and activity.^134-136 NK cell infusions administered in the peritransplant period have been shown to be safe and well-tolerated in early-phase clinical trials.^137-139 Additionally, Choi et al¹³⁷ noted that donor-derived NK cell infusion after reduced-intensity conditioning haplo-HCT was associated with reduced relapse and no increase in GVHD rates when compared with historical controls. Since a limitation of NK cells is their relatively low frequency in grafts and their short in vivo persistence, Ciurea et al¹⁴⁰ used feeder cells expressing membrane-bound IL-21 to expand NK cells ex vivo and reported a low incidence of relapse with limited GVHD upon infusion into patients before and after haplo-HCT. In a similar approach, cytokine-induced memory-like NK cells from haploidentical donors can be generated by ex vivo culture with IL-12, IL-15, and IL-18 and exhibit enhanced antitumor activity.¹⁴¹ Clinical trials using these cells as a prophylactic DLI and for treatment of frank relapse are ongoing (ClinicalTrials.gov identifier: NCT02782546 and NCT03068819). To expand NK cells in vivo, Romee et al¹⁴² administered ALT-803, an IL-15 superagonist, in a phase I study of 33 patients relapsing > 60 days post-HCT and reported the activation and proliferation of CD8+ T and NK cells with an overall response rate of 19%.

Direct engineering of T cells to target AML cells by transduction with T cell receptors previously identified to recognize leukemia-associated antigens (TCR transgenic T cells) or with chimeric antigen receptors (CAR-T cells) may provide robust AML cell killing after HCT. With a median follow-up of 44 months, Chapuis et al¹⁴³ recently reported no relapse in 12 patients with high-risk, heavily pretreated AML treated with WT1-specific TCR transgenic CD8+ T cells after transplantation. Currently, no trials of CAR-T cells for AML in the post-HCT relapse setting have been reported, but the reports of anti-CD19 CAR-T cells used after HCT suggest that this approach may be effective and well-tolerated with a low incidence of GVHD.^144,145

Bispecific Antibodies

Bispecific antibodies work by engaging tumor cells with immune effector cells and directly activating immune cells independently of MHC/TCR interactions. In acute lymphoblastic leukemia (ALL), treatment with the bispecific CD19xCD3 agent blinatumomab has been shown to be safe and effective, including after HCT.¹⁴⁶ In AML, several early-phase clinical studies are testing bispecific antibodies targeting CD33, CD123, and CLL-1 in patients with R/R disease, including patients who relapsed after HCT.¹⁴⁷ A potential advantage of bispecific antibodies in the post-HCT setting is that since they do not rely on MHC/TCR interactions, they would be predicted to work in cases where HLA expression has been lost after relapse post-HCT. Indeed, Rovatti et al¹⁴⁸ showed in a preclinical model that an anti-CD33/CD3 bispecific antibody was able to restore T cell activation by AML cells that had lost their mismatched HLA haplotype after haploidentical transplantation. In addition, since T cell engagement leads to IFNg release within the tumor microenvironment, it is possible that the IFNg-induced restoration of MHC class II expression on AML cells could contribute to bystander immune cell killing. We have generated data using both an anti-CD3/CD123 bispecific molecule (flotetuzumab, MGD006) and anti-CD123 CAR-T cells that suggest that targeting AML cells in this fashion leads to the upregulation of MHC class II expression on surrounding AML blasts (unpublished data, see Fig 1). This reinduction of HLA class II expression can potentially restore the GVM effect and effectively treat a subset of relapsed AML post-HCT.

FIG 1.

T cell immunotherapy can upregulate MHC class II expression on AML blasts. (A) In 30%-50% of AML patients relapsing after allogeneic hematopoietic cell transplantation, AML blasts have a decrease or loss of surface MHC class II expression, impairing the ability of miHA- and neoantigen-specific T cells to recognize AML blasts and exert a GVM effect.^27,38 (B) Introduction of a bispecific antibody or genetically engineered T cell (ie, CAR T cell) capable of recognizing LAAs on AML blasts will lead to non-MHC–restricted T cell recognition of the AML blasts. (C) Non-MHC–restricted T cell recognition of AML blasts will activate the T cells and stimulate the release of IFNg. (D) The release of IFNg will upregulate MHC class II surface expression on surrounding AML blasts, including both blasts that are recognized and unrecognized by the T cell immunotherapy. (E) Upregulation of MHC class II expression on AML blasts will allow for miHA- and neoantigen-specific T cells to recognize and kill the AML blasts, leading to a GVM effect. AML, acute myelogenous leukemia; CAR-T cell, chimeric antigen receptor T cell; GVM, graft-versus-malignancy; IFNg, interferon gamma; IFNgR, interferon gamma receptor; LAA, leukemia-associated antigen; MHC II, major histocompatibility antigen class II molecule; miHA, minor histocompatibility antigen; scFv, single-chain variable fragment.

Gemtuzumab Ozogamicin

Several case reports and small trials testing gemtuzumab ozogamicin (GO) monotherapy and combination therapy in maintenance or salvage settings suggest that the drug exhibits clinical activity after HCT, but some concerns of myelosuppression and hepatic veno-occlusive disease were noted.¹⁴⁹ Recently, Genthon et al¹⁵⁰ reported an overall response rate of 72% (13/18) with a 1-year OS of 54% in patients with AML relapsing after HCT who were treated with GO in combination with intensive chemotherapy. However, all patients experienced grade III-IV neutropenia and thrombocytopenia. Ongoing clinical trials are testing fractionated dosing of GO alone to treat measurable residual disease (ClinicalTrials.gov identifier: NCT03737955) or GO in combination with CPX-351 chemotherapy to treat relapse (ClinicalTrials.gov identifier: NCT03904251).

DISCUSSION

Relapse after HCT portends a very poor prognosis, and current management approaches such as DLI and second HCT have a modest response rate and significant risk of toxicity. Recent research has identified genetic and epigenetic changes that have resulted in the downregulation of HLA molecules and upregulation of inhibitory checkpoint molecules. These changes suggest a model wherein loss of immune effector cell function contributes to relapse and suggests possible approaches for preventing or treating relapsed malignancies that exploit the GvL effect after HCT.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Financial support: John F. DiPersio

Administrative support: John F. DiPersio

Provision of study materials or patients: John F. DiPersio

Collection and assembly of data: Joseph C. Rimando, John F. DiPersio, Matthew J. Christopher, Michael P. Rettig

Data analysis and interpretation: Joseph C. Rimando, Michael P. Rettig, John F. DiPersio, Matthew J. Christopher

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Biology of Disease Relapse in Myeloid Disease: Implication for Strategies to Prevent and Treat Disease Relapse After Stem-Cell Transplantation

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I =Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).

Joseph C. Rimando

Stock and Other Ownership Interests: Merck

Michael P. Rettig

Consulting or Advisory Role: Rivervest

Research Funding: MacroGenics, Amphivena Therapeutics, Maxcyte, BiolineRx

Patents, Royalties, Other Intellectual Property: MPR has pending patent applications (PCT/US2017/059777, integrin inhibitors and chemokine receptor agents; PCT/US2017/059733, integrin antagonists) and reports royalties received from the patent applications during the conduct of the study

John F. DiPersio

Stock and Other Ownership Interests: Magenta Therapeutics, WUGEN

Honoraria: Incyte

Consulting or Advisory Role: Cellworks, Rivervest, Magenta Therapeutics, Incyte

Research Funding: Amphivena Therapeutics, Macrogenics, Incyte, WUGEN, BiolineRx, Maxcyte, Bigelow Aerospace

Patents, Royalties, Other Intellectual Property: CD7 and CD2 Knockout for CART to CD7 and CDL, Duvelisib for treatment of cytokine release syndrome (CRS), NT-17 to enhance CART Survival, Novel WU mobilizing compounds, Selection of IMPDH Mutant Stem Cells, IFNg, upregulate MHCII for relapsed AML, Dextran based molecules to detect CAR-T cells, Combining integrin inhibitor with chemokine binders, 016131, JAK and calcineurin inhibition, solid organ transplant, VLA4, gro-b, Triple Combination - CXCR2, VLA-4, gro-b, Targeting IFNR/CSCR3 in GVHD, WU/SLU compounds VLA4 and CXCR2

Travel, Accommodations, Expenses: Incyte, Macrogenics, Magenta Therapeutics

No other potential conflicts of interest were reported.