AbstractAbstract 2026 Background:In recent years stem cell transplantation (SCT) has been used increasingly for the treatment of different oncohematological diseases. Although improvements have been obtained, bacterial, viral and fungal infections continue to represent the major challenge. SCT in MM increases patient’s vulnerability to infections due to immunological changes related to the conditioning and the disease itself. The increased cytokine expression, that involve the pathogeneses of multiple myeloma, like TNF-α, IL-6 and 10, have an important relationship with the immune response and disease progression. Single Nucleotide Polymorphisms (SNPs) in the promoter region of cytokine genes are responsible for altering the levels of expression, thereby affecting the evolution of the disease and the immune response. It has been considered that the immune response is modulated by the interaction of several factors or gene products on related metabolic pathways. We have hypothesized that SNPs in the IL-6, IL-10 and TNF-αgenes may be associated with the susceptibility to infection, either when considered alone or combined in haplotypes. More importantly, however, is the gene-gene interaction analysis of the SNPs in these biologically related genes. Objectives:We performed association analysis of single polymorphisms, the combination of SNPs in haplotypes and epistasis or gene-gene interaction analysis. Methods:We have clinically assessed the infection status of the patients for the following conditions: fever of unknown origin, death during the neutropenic phase, fungeous blood stream infection (F.BSI), superinfection, bacteremia (Gram-positive BSI and Gram-negative BSI), invasive fungeous infection. Genomic DNA was extracted from mobilized peripheral-blood stem cells of 148 patients. Genotyping was carried out for the SNPs IL-6 (-174G/C), IL-10 (-1082A/G, -592A/C e -819T/C) and TNF-α (-308G/A) using the Real-time PCR assay and PCR followed by restriction length fragment polymorphism (PCR-RFLP). Haplotype frequencies were estimated with Haplo.stats® and all the other statistical analysis were performed with SPSS®. Gene-gene interaction analysis was performed by multifactor dimensionality reduction (MDR®) method. Results:We have not observed deviations from the Hardy-Weinberg equilibrium for the genotypes. The TNF-α polymorphism was significantly associated with fever of unknown origin (p=0.03), IL-6 polymorphism was associated with fungemia (p=0.04) and death during neutropenia (p=0.02), and IL-10 polymorphism was associated with superinfection (p=0.02). Interestingly, the IL-10 “ATA” haplotype was also associated with superinfection (p=0.016). In addition, the “ATA” haplotype was associated with F.BSI (p=0.004) and the “ACC” haplotype was associated with Gram-negative bacteremia (p=0.040) and superinfection (p=0.048). Epistasis analysis revealed a significant interaction model between TNF-α G308A and IL-6 G174C polymorphisms when compared the groups with and without bacteremia (Balanced Accuracy = 0.6021, Cross Validation Consistency = 10/10, p=0.0380). Conclusions:The results showed that despite the primary predisposition to infection typical of multiple myeloma, the presence of the variants studied significantly affected the susceptibility to serious infections and outcomes of ASCT. Moreover, our findings from epistasis analysis indicate that the interaction of polymorphisms may be more important than the effects of single polymorphisms for the immune response associated with the infection susceptibility in ASCT. Disclosures:No relevant conflicts of interest to declare.