Monika M. Kutyna

Infection is a major cause of morbidity and mortality in myelodysplastic syndromes (MDS), yet infection risk remains incompletely defined in the contemporary treatment era. We conducted a retrospective study of 708 patients with MDS to characterize the incidence, microbiology, temporal dynamics, and predictors of infection-related hospitalization over two decades. Overall, 78.8% (n=558) of patients required hospitalization, of which 69.9% were infection related. Infection-related hospitalization was independently associated with inferior overall survival. In multivariable Cox proportional hazards model, comorbidity burden, red blood cell transfusion dependence, higher IPSS-R risk, exposure to chemotherapy or stem cell transplantation, and severe neutropenia independently predicted infection. Among azacitidine-treated patients, 71.4% experienced infection-related hospitalization, with nearly three-quarters occurring within the first six cycles, identifying a critical early vulnerability window. Neutropenia remained the dominant driver of infection risk; however, immune dysfunction independently increased susceptibility. Low monocyte counts, cytokine dysregulation, and TP53 mutations identified high-risk patients despite preserved neutrophil counts, reflecting impaired myeloid reserve. Consistent with this dynamic vulnerability, recent infection or severe neutropenia (<0.5×10⁹/L) increased subsequent infection risk by 2.47-fold. We observed a concerning shift in antimicrobial resistance, with 14% of Gram-negative infections producing extended-spectrum beta-lactamases, 21% of Pseudomonas aeruginosa isolates resistant to piperacillin-tazobactam, and vancomycin-resistant enterococci prevalence increasing from 14% to 46%. Together, these findings identify infection in MDS as a dynamic, prognostically important complication driven by cytopenia, treatment, and immune dysfunction. This supports time-adapted risk stratification, targeted prevention, and antimicrobial stewardship, particularly during early treatment and high-risk disease phase.

Supplementary Table 5 shows distribution of somatic mutations across different studies in therapy-related acute myeloid leukemia (t-AML) and primary AML

Supplementary Table 7 shows allogeneic stem cell transplant outcome across different studies in t-MN

Supplementary Table 6 shows distribution of germline mutations across different studies in t-MN

Supplementary Table 1 shows distribution of t-MN cases according to type of primary cancer/diseases

Supplementary Table 4 shows distribution of somatic mutations across different studies in therapy-related MDS (t-MDS) and primary MDS

Supplementary Table 2 shows comparative frequency of cytogenetic abnomalities. (A) p-MN and t-MN; (B) p-MDS and t-MN stratified according to IPSS-R; (C) p-AML and t-AML stratified according to MRC/ELN.

Supplementary Table 3 shows distribution of somatic mutations across different studies in therapy-related myeloid neoplasms (t-MN) and primary MN

Table S1 showing AML primary cell details for metabolomics

Supplementary Figure file showing Figures S1-S7