Abstract 2344

Introduction:

Allogeneic hematopoietic cell transplantation (HCT) can cure hematologic malignancies through beneficial graft versus leukemia (GVL) allo-immune responses, but is limited by graft versus host disease (GVHD). Studies have shown allogeneic antibodies (allo-Ab) develop against minor histocompatibility antigens encoded on the Y-chromosome after sex-mismatch HCT in association with chronic GVHD (cGVHD) and persistent disease remission. Recombinant human protein microarrays facilitate multiplex antibody quantification and have successfully identified GVL antigens in single patient studies. This study applies protein microarray technology to compare antibody development in 43 leukemia patients in relation to allogeneic HCT outcomes. We hypothesized allo-Ab quantification may identify patients with cGVHD and identify unique GVL antigens in acute myeloid or lymphoblastic leukemia.

Method:

We enrolled a consecutive series of acute leukemia patients undergoing HLA-identical myeloablative allogeneic HCT and studied the 20 AML and 23 ALL patients who survived at least one year after transplantation. Patients received FK506 and methotrexate immune prophylaxis. Only 5 of 43 grafts were bone marrow and 33% of donors were unrelated. Extensive cGVHD developed in 24 of 43 (56%) patients. Plasma samples collected pre-HCT, one year post-transplant, and from 45 healthy donors were analyzed for antibodies using v4.1 ProtoArrays (Invitrogen, Carlsbad, CA). These are nitrocellulose coated glass slides printed in duplicate with 8350 unique human proteins epitope-tagged with GST. Plasma diluted 1:500 was incubated 1.5 hours and bound antibodies were detected using fluorochrome conjugated anti-human IgG. Slide processing and fluorochrome detection were normalized by comparison to known microarray spotted IgG concentrations. Overall, mean fluorescence intensity (MFI) of duplicate spots showed excellent agreement (R2=0.95). Each antibody response was normalized for target protein concentration using anti-GST measurements. Antibody responses were reported as Z scores (number of standard deviations above each sample's average MFI). For each sample, the sum of Z scores>2 (SZ2) was analyzed in relation to clinical outcome.

Results:

The SZ2 measured one year post-HCT ranged from 279 to 6652 and did not correlate with total IgG (R2=0.26). Univariate analysis of transplant risk factors and clinical outcomes showed SZ2 associated with extensive chronic GVHD. Using a SZ2 cutoff of 3700, the relative risk of developing extensive cGVHD was 1.9 (95%CI 1.14–3.07) with a positive predictive value of 84.2%. These allo-Ab (SZ2>3700) developed in extensive cGVHD patients despite elevated corticosteroid therapy (p<0.001). Further, a set of 25 proteins were recognized by allo-Ab in more than 50% of cGVHD patients including fibroblast growth factor 12. Strong allo-Ab responses developed against FGF12 in 16/24 (66%) of extensive cGVHD patients (Z scores ranging 3–22, median of 8).

With a median post-HCT follow-up of 2.7 years, only 6 patients have relapsed and SZ2 did not associate. In order to identify AML and ALL specific GVL antigens, we screened for allo-Ab targets exclusively recognized in either AML or ALL patients but not detected in 45 normal donors. Twelve AML and 20 ALL candidate GVL targets were identified in at least 6 of 20 (30%) patients including Myosin Light Chain Kinase 2 (MYLK2) in AML patients. These candidate GVL antigens are currently being validated using banked samples collected after allo-HCT from patients with a variety of malignancies.

Conclusion:

Protein microarray technology and normalization methods enable quantitative antibody measurements across complex patient populations. Here we show elevated allogeneic antibodies measured by protein microarray one year after transplant associated with extensive cGVHD development. We conclude protein microarrays can identify GVHD antibody signatures and that total allo-Ab quantification can be monitored in real-time for cGVHD development aiding immune suppression management.

Disclosures:

Miklos: Novartis: Honoraria, Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution