AML relapse after transplant is common, associated with dismal out comes, and remains a major obstacle to HSCT success. Up to 40% of patients with AML/MDS relapse after undergoing HLA-matched HSCT. Powerful donor-derived immune responses underlie the GVL effect, and immune escape (i.e., tumor evasion of the donor immune system) contributes to relapse after HSCT.
In HLA-matched HSCT, presentation of host hematopoietic and/ or tumor epitopes within the concordant HLA context can invoke T cell receptor (TCR)-mediated recognition and elimination by donor T effector (Teff) cells to induce curative GVT responses that may be antagonized by loss of Class I HLA or suppression of Class II HLA on tumor cells, by suppression of co-stimulatory immune pathways, by immune suppressive T regulatory (Treg) cells, or by activation of inhibitory immune checkpoint pathways. While the loss of tumor Class I HLA (typically mismatched allele) can occur during relapse after HLA-mismatched and haploidentical HSCT, it is uncommon in HLA-matched and absent in UCB HSCT, highlighting the importance of individualized tumor versus donor immune interactions in each clinical context.
Inhibitory immune checkpoint pathways play an important role in relapse after HSCT. The engagement of cytotoxic T-lymphocyte associated protein 4 (CTLA-4) and programmed death 1 (PD-1) receptors by their ligands B7-1/2 and PD-L1/2, respectively, inhibits effector T cell function. Tumor cells often express these inhibitory ligands, thereby selectively blocking antitumor immunity. In murine models, PD-1 blockade led to an increase in GVHD, whereas selective CTLA-4 blockade with Ipilimumab (Ipi) to treat late HSCT relapses augmented GVT without accelerating GVHD. Outside of HSCT, Ipi monotherapy has limited efficacy in relapsed AML/ MDS.
For patients with targetable mutations, or for therapies broadly augmenting GVL and controlling disease relapse, maintenance strategies to prevent AML relapse in the post-transplant context are worth evaluating. For instance, early-phase trials of maintenance therapy with HMAs for MDS/AML (e.g., azacytidine), and targeting of FLT3-ITD AML (e.g., sorafenib) reported preliminary safety and efficacy. In the randomized phase II SORMAIN study, FLT3-ITD AML patients who had undergone HSCT and were stably engrafted without grade II or higher acute GVHD were randomly assigned to sorafenib maintenance for up to 24 months versus placebo. At 30 months, OS favored the sorafenib arm, with a HR of 0.447 (P = .03). Preliminary results of the randomized phase II RADIUS trial of post-transplant midostaurin likely support this conclusion, with a 43% reduction in the risk of FLT3-ITD AML relapse. Interestingly, recent data show that sorafenib also promotes GVL activity in mice and humans through IL-15 production in FLT3-ITD-mutant leukemia cells.
While falling donor chimerism after HSCT can identify cohorts at higher relapse risk, more specific and sensitive MRD assessments in the post-HSCT context can better predict relapse. Importantly, early identification of patients at high risk of relapse after allogeneic HSCT offers the opportunity for targeted early intervention prior to morphologic relapse. Prospective prevention trials focused on the high-risk setting offer an informative population for evaluation of intervention efficacy and are a high priority. The ongoing randomized phase III trial of gilteritinib (BMTCTN 1506, NCT02997202) will help definitively address the issue of FLT3-ITD maintenance post-transplant, especially in those at high risk, as a highly sensitive and specific MRD assay will be used throughout the trial to provide information about which patients may or may not benefit from maintenance therapy. Similar trials for the prevention of AML relapse after transplant in patients with other targetable mutations (e.g., IDH1/2) are eagerly anticipated.
The treatment options for patients with morphologic disease relapse after HSCT include withdrawal of IS, DLI, chemotherapy (including novel agents), second allogeneic transplantation, or supportive care. Taper of IS and DLI are used especially in patients without GVHD to enhance antileukemic GVL effect. DLI has limited efficacy, in the range of 15% to 30% for patients with relapsed AML, and many responses are incomplete or temporary. Of note, recommended DLI dose is ∼1-log lower in haplo-HSCT; and it is not expected to have efficacy in relapse associated with mismatched HLA loss on tumor cells but still carries a risk of GVHD and should be avoided in that context. DLI outcomes are better in patients who relapsed beyond 6 months after HSCT, those with lower leukemic burden (less than 35% BM blasts), patients with good-risk cytogenetics, and those who achieved remission before DLI.
Novel options to enhance curative GVL effect include infusion of selected donor lymphocyte subpopulations via mDLI (e.g., CD25 Treg-depleted DLI), other immune effectors (e.g., cytokine-induced memory-like [CIML] NK cells), and use of immune checkpoint inhibitors (e.g., CTLA-4 blockade) A second allogeneic HSCT has demonstrable but limited efficacy, and the duration of remission of more than a year after initial HSCT is the major predictor of survival.
