A new variable to study in transplant: APOE variant status

In this issue of Blood, Ronnacker et al¹ report the results of a retrospective analysis of associations between Apolipoprotein E (APOE) gene variants and clinical outcomes in patients undergoing allogeneic hematopoietic cell transplantation (allo-HCT) for acute myeloid leukemia (AML). This highly original study of over 700 patients presents compelling data that indeed APOE variant status is a variable that correlates with posttransplant toxicities and outcomes (see figure). Specifically, the authors found that transplant recipients with an APOE2 allele had a higher risk of chronic graft-versus-host disease (GVHD) and nonrelapse mortality (NRM) compared with recipients carrying the APOE4 variant. Moreover, for recipients without APOE2, the presence of APOE2 in the donor was associated with increased risk of acute GVHD compared with non-APOE2 donors. Given that APOE2 is a common variant in the general population, these findings, if validated in a prospective study, present a new variable to consider in risk assessment for transplant and optimal donor selection.

APOE, a glycoprotein originally identified in the context of lipid transport,² is a molecule now associated with a myriad of effects related to metabolism, immunomodulation, and neurodegenerative pathology.^3-5 The 3 most prevalent germ line variants of APOE in humans are APOE2, APOE3, and APOE4, with estimated frequencies of approximately 10% to 15%, 60% to 70%, and 15% to 20%, respectively.⁵ Notably, although these allelic variants differ by 1 or 2 amino acids, these differences lead to changes in the structure and function of APOE.^6,7 The list of conditions with connections to APOE variant status continues to grow. For instance, although APOE4 is a well-established risk factor for Alzheimer disease,^4,APOE2 is considered relatively protective⁵; in contrast, APOE4 has been linked to improved outcomes in melanoma compared to patients with APOE2.⁸ 

Across neurodegenerative, inflammatory, and malignant diseases, there appears to be a strong link between disease pathology and aberrant immunity.^4,9 Ronnacker et al therefore set out to investigate a very fundamental question: does APOE variant status modulate outcomes in allo-HCT, the only established long-standing immunotherapy for AML? To do this, they performed APOE genotyping in a discovery cohort of >300 patients (in a single center in Germany) and validation cohort of >400 patients (in multiple centers in Germany) who underwent allo-HCT for AML in first remission. In parallel, they performed donor APOE genotyping when possible, from >300 patients and >150 patients in each cohort, respectively. The authors then measured overall survival stratified by APOE variant status, revealing that the presence of an APOE2 allele in transplant recipients was associated with inferior posttransplant survival in both cohorts. In contrast, APOE2 variant status did not predict survival in patients with AML studied from alternative patient cohorts that did not undergo transplantation. Further analyses to understand these distinct transplant outcomes revealed that recipients carrying APOE2 had a higher incidence of chronic GVHD and NRM, effects that remained significant after adjusting for multiple other critical transplant variables.

Having established the deleterious effects of the APOE2 allele in the transplant recipient, the authors explored the effect of APOE variant status in the transplant donor. Again, presence of the APOE2 allele in a donor paired with a recipient without an APOE2 allele conferred worse outcomes, specifically related to acute and chronic GVHD. Taken together, these findings suggest that presence of an APOE2 allele in either an allo-HCT recipient or donor in specific scenarios may negatively affect GVHD-free, relapse-free survival.

These findings motivate mechanistic studies to elucidate how APOE variant status shapes posttransplant immunity and ultimately outcomes. The authors investigated reconstitution of major immune subsets posttransplant (T cells, B cells, and natural killer cells), but no clear differences stood out from this initial analysis. Given the increased risk of GVHD in the context of the APOE2 variant, one plausible explanation is that peptides derived from ApoE2 may function as a minor alloantigen, thus leading to deleterious alloimmunity from donor T cells not tolerant to this mismatched peptide. To explore this hypothesis, the authors conducted an in silico analysis to predict the HLA-binding affinities of peptides derived from ApoE2. Only 3 HLA alleles (B∗08:01, B∗15:01, and B∗51:01) stood out from this analysis and the authors did not identify significant differences in GVHD incidence in patients with a donor-recipient APOE2 mismatch with respect to having one of these identified alleles. Furthermore, the authors specifically looked for differences in GVHD risk in transplants between donors and recipients mismatched on APOE2 status (APOE2-negative donor to APOE2-positive recipients) but did not find increases in GVHD incidence. Therefore, additional rigorous investigation of innate and adaptive immune subsets will be essential to understand what is driving this intriguing clinical finding.

This study does have several limitations to consider, including (1) its retrospective nature, (2) its focus on only patients with AML in remission, (3) the fact that the majority of patients received HLA-matched peripheral blood stem cell grafts, and (4) the inclusion of only patients from a single country. The potential effects of APOE variant status on transplants from alternative donor sources, for high-risk groups related to disease or comorbidity status, as well as for nonmalignant conditions warrant further investigation. Nonetheless, this compelling study provides the first evidence associating APOE2 variant status with adverse outcomes following allo-HCT, and therefore APOE variant status may ultimately be a key variable to consider when estimating GVHD and NRM risk for allo-HCT. Further validation via a prospective study paired with deep mechanistic analyses is essential to confirm these very promising results and discover biological underpinnings.

Conflict-of-interest disclosure: S.D. reports research funding from The National Institutes of Health (NIH)/National Cancer Institute (NCI) KO8 (1K08CA29327-01), the Memorial Sloan Kettering (MSK) Cancer Center Leukemia Specialized Program of Research Excellence (SPORE) Developmental Research Program (NIH/NCI P50 CA254838-01), Parker Institute for Cancer Immunotherapy, the MSK Center for Tumor-Immune Systems Biology Pilot, and the MSK Gerstner Physician Scholar Program. S.K. declares no competing financial interests.