Abstract / Summary
Abstract Background Infection during chemotherapy-associated neutropenia is a major clinical problem in children with acute leukemia. Rapid host-response biomarkers may provide complementary information while microbiological results are pending. This study characterized IL-6/IL-10 response patterns and their associations with lymphocyte subsets and microbiological Gram classification using the reported study dataset. Methods This single-center retrospective study reviewed 268 clinical episodes recorded between September 2022 and March 2026. A total of 127 infection episodes met the original study criteria. Episodes were described using two neutral cytokine phenotypes: concomitant elevation of IL-6 and IL-10 (dual-elevation profile) and IL-6 elevation with absent or minimal IL-10 elevation (IL-6-predominant profile). Cytokines and lymphocyte subsets were compared between profiles. Concordance with blood culture, sputum culture, and next-generation sequencing findings was examined in microbiologically documented bacterial episodes. Associations with response to meropenem or vancomycin were explored secondarily. Results Among 62 neutropenic infection episodes, 32 had a dual-elevation profile and 30 had an IL-6-predominant profile. The dual-elevation group had higher absolute lymphocyte, CD3 + T-cell, CD4 + T-cell, CD8 + T-cell, B-cell, and IL-10 levels (all P < 0.05). Similar differences in absolute lymphocyte, CD3+, CD4+, CD8+, B-cell, and IL-10 measures were observed in 65 non-neutropenic episodes. Among 24 microbiologically documented bacterial episodes, 12/14 (85.71%) Gram-negative episodes showed the dual-elevation profile, whereas 9/10 (90.00%) Gram-positive episodes showed the IL-6-predominant profile (P = 0.001). In the exploratory treatment-response subset, 27/32 (84.38%) meropenem-responsive episodes showed the dual-elevation profile and 28/31 (90.32%) vancomycin-responsive episodes showed the IL-6-predominant profile (P < 0.001). Conclusions Distinct IL-6/IL-10 response patterns were associated with different lymphocyte-subset profiles and showed substantial concordance with microbiological Gram classification in this retrospective cohort. Because cytokine information contributed to the original phenotype definition, the findings should be considered hypothesis-generating and require prospective validation using independently defined microbiological outcomes.