A phenotypic hallmark of the myelodysplastic syndromes (MDS) is ineffective hematopoiesis, resulting in anemia, neutropenia, and thrombocytopenia. The extent to which the stromal microenvironment interacts with MDS progenitors to influence the natural history of the disease is unknown. A number of proinflammatory cytokines, including TNFα, and IL-6, are increased in MDS patients versus normal subjects. In addition, the TNF and IL-6 genes have functional single nucleotide polymorphisms (SNPs) that are linked to cytokine hypersecretion. However, the influence of these SNPs on the MDS phenotype remains poorly defined. Our central hypothesis is that both cell autonomous and cell non-autonomous factors cooperate to define the MDS phenotype. We postulate that underlying host factors do not predispose to the development of MDS but act as genetic modifiers of disease severity, particularly with regard to the degree of cytopenias, severity of infectious complications, and disease progression. In the current study, we analyzed the frequency of the TNFα -308G/A and -238G/A, IL-6 -174G/C, and macrophage migration inhibitory factor (MIF) -173G/C and -794 CATT repeat promoter region polymorphisms and determined whether these SNPs were associated with adverse clinical and laboratory outcomes in a large cohort of MDS patients. Genotype analysis on DNA samples from 328 MDS patients was performed and the respective TNF, IL-6 and MIF haplotype frequencies were correlated with the following outcomes: age, IPSS score, karyotype, blast percentage, FAB classification, survival, anemia, neutropenia, and thrombocytopenia. As shown in Table 1, multivariate analysis of the TNFα -308G/A SNP revealed that the AA genotype (high-expressing) was significantly associated with an abnormally low ANC (<1500) as well as frank neutropenia (ANC<1000) (p<0.05 for each comparison, Cox regression).GenotypeANC<1500ANC<1000AA5/5 (100%)*4/5 (80%)*GA36/104 (35%)22/104 (21%)GG86/219 (39%)57/219 (26%)Table 1. TNF-308A/A genotype is associated with neutropenia in MDS. *p<0.05 for each ANC level analyzed.As shown in Table 2, multivariate analysis of the TNFα -238G/A SNP revealed that the presence of a high-expressing genotype (AA + GA) was significantly associated with NCI Grade II or greater anemia (Hgb<10g/dl) (p<0.05, Cox regression). After controlling for the presence of red blood cell transfusions, the frequency of the A allele in patients with Hgb<10g/dl remained statistically significant (p<0.05). Furthermore, whereas the A allele was present in 14% of anemic individuals (Hgb<12.5g/dl, n=41), no patients with a normal Hgb (n=26) carried the A allele (p=0.03, Chi square). This also remained significant after controlling for red cell transfusions (p=0.02).Table 2. The presence of an A allele in the TNF-238G/A SNP is associated with NCI Grade II anemia in MDS. *p<0.05.GenotypeHgb<10g/dlGA + AA26/41 (63%)*GG136/287 (47%)In contrast, the frequencies of the IL-6 or MIF SNPs did not have a statistically significant effect on any of the measured outcomes. To evaluate the effect of TNFα on erythroid colony formation, human peripheral blood CD34+ cells from healthy adults were cultured in semi-solid media in the presence of IL-3 (10ng/ml), erythropoietin (1U/ml), SCF (50ng/ml) and increasing doses of TNFα (0–100ng/l) and erythroid burst-forming units (BFU-E) measured on days 7 and 14. In the presence of 25 and 100ng/ml TNFα, the percentage of BFU-E colonies was reduced by 50%. Furthermore, the erythroid colony suppressing affects of TNFα were reversed to baseline in a dose-dependent manner with the addition of the anti-TNF agent infliximab. These data indicate that the presence of high-expressing polymorphic alleles in the TNFα gene promoter is independently associated with both neutropenia and the severity of anemia in a large cohort of MDS patients. TNFα-overexpression may function as a cell non-autonomous factor that interacts synergistically with intrinsic defects within MDS progenitors to promote hypoplasia within the hematopoietic lineage, potentially increasing the severity of cytopenias in patients with MDS.