Abstract 3481

A significant proportion of patients with idiopathic AA respond to immunosuppressive therapy (IST), suggestive an ex juvantibus autoimmune pathogenesis. Refractory cases may be due to either insufficient intensity of IST or exhaustion of stem cell reserves. Similarly, there may be a non-immunological component of the disease that cannot be distinguished clinically to date. Multiple trials of various ISTs often delay potentially curative therapy with bone marrow transplantation (BMT); hence, identification of this subtype is of great importance.

Immune-mediated AA may have measurable immunogenetic determinants, such as HLA genes, that modify susceptibility to, character of, or intensity of autoimmune reactions. We hypothesize that genetic polymorphisms may exist in immune cytokines, cytokine receptors, and immune regulatory genes that may be markers of immune-mediated disease.

We applied a custom cancer chip (Illumina) containing 211,155 probes for mostly non-synonymous single-nucleotide polymorphisms (SNPs) to perform focused bioanalysis on 32 immunogenetic polymorphisms. Our cohort included 152 patients; AA (N=91), AA/PNH (N=38), and PNH (N=23). Results were compared to internal control and previously published results with comparable populations. Median age was 45 years (5–80 years); 92% (N=109) had normal cytogenetics; 67% (N=96) had IST with an overall response rate (ORR) of 65%; 11% (N=17) underwent hematopoietic stem cell transplant; 43% (N=49/114) were positive for HLA DR15. The following genotypes had higher incidence in our AA cohort vs. controls: IL1A C(1202)T (rs1800794) CC 60% vs. 35% (p=<.001), IL4 –33 T>C (rs2070874) CC 68% vs. 29% (p=<.001), IL4 –590 C>T (rs2243250) CC 95% vs. 74% (p=<.001), IL2 T-330-G (rs2069762) TT 47% vs. 37% (p=.04), IL6 NT565A>G (rs1800797) AA 17% vs. 8% (p=.006), IL10 –1082G>A (rs1800896) AA 39% vs. 20% (p=<.001), TNF -308 G>A (rs1800629) GG 95% vs. 71% (p=<.001) and IL10 –592 A>C (rs1800872) AA 16% vs. 6% (p=.005). Among AA subtypes, we compared the frequency of each genotype: IL10 –592 A>C (rs1800872) was more frequent in AA vs. PNH (p=.003); both homozygous genotypes, CC and TT of TGF b1 codon 10 +869 C/T (rs1800470) were increased in AA vs. AA/PNH (p=.007). Frequency of CC genotype of CR2 C>T (rs2802221) was increased in AA vs. PNH (p=.003). Differences were also present in CR2 (S663P exon 11, rs4308977) and CR2 (rs6667140) in AA vs. AA/PNH (p=.001), AA vs. PNH (p=.031), and in AA vs. AA/PNH (p=.002). Next, we identified prognostic factors for response via univariate analysis which showed a trend towards higher ORR in AA/PNH vs. AA (88% vs. 59%, p=.06), as previously described. However, neither karyotype nor presence of HLA DR15 affected ORR. Interestingly, there was a trend towards higher ORR in patients with the following genotypes: IL4 –33 T>C CC vs. CT vs. TT (80% vs. 53% vs. 50%, respectively), p=.08; IL4 –590 C>T CC vs. CT vs. TT (69% vs. 33.3% vs. 0, respectively), p=.06. We analyzed (logistic regression) 3 variables (presence of PNH, IL4 –33 T>C SNP, and IL4 –590 C>T SNP) identified in univariate analysis (p<.15) and found that polymorphism in IL4 –33 T>C particularly the CC genotype is prognostic for response (p=.038) while IL4 –590 C>T polymorphism showed a trend towards significance (p=.09).

We also employed unbiased screening of non-synonymous SNPs in 116 cases and confirmed in an additional 120. 3 SNPs were outstanding based on differential frequency between patients and controls. There was a difference between dominant/minor (C/T) alleles rs13050238 (CRYZL1) between patients and controls [213/9 (95%, dominant allele) vs. 3747/25 (99%), p<1×108] suggesting carriers of at least one copy of the T allele (CT/TT) are at higher risk of developing AA. Similarly, the minor allele G (A/G) in rs6451268 (RANBP3L) has protective potential [194/30 (87%) vs. 3721/207 (95%), p<1×107]. Minor allele T (C/T) in rs911178 (ZKSCAN3) may also be a protective allele [193/33 (85.53%) vs. 3641/287 (92.69%), p=<1×105]. This gene may regulate expression of genes favoring tumor progression in colorectal cancer.

In summary, the T allele of IL1A 1202, IL4 –33, IL4–590; G allele of IL2 –330, IL6 NT565, IL10 –1082; A allele of TNF -308 and C allele of IL10 –592 may be protective against development of AA. SNPs in IL4 –33 T>C and IL4 –590 C>T may be predictive of response to IST and may be considered along with the presence of PNH as favorable prognostic factor.


Maciejewski:NIH: Research Funding; Aplastic Anemia&MDS International Foundation: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.