TIM-3 Expression Level on AML Blasts Correlates With Presence of Core Binding Factor Translocations Rather Than Clinical Outcomes

BACKGROUND: T-cell immunoglobulin and mucin domain-containing molecule 3 (TIM-3) expresses on leukemic stem and progenitor populations of non-M3 acute myeloid leukemia (AML) as well as T lymphocytes. TIM-3 is thought to be involved in the self-renewal of leukemic stem cells and the immune escape of AML cells, however its correlation with AML prognosis is still controversial and worthy of further investigation. METHODS: we simultaneously assessed TIM-3 expression levels of leukemic blasts and T lymphocytes in the bone marrow of de novo AML patients using flow cytometry. The correlations of TIM-3 expression between leukemic blasts and T lymphocytes and the correlations of TIM-3 expression with various patient parameters were analyzed. In addition, the Cancer Genome Atlas (TCGA) data of AML patients were acquired and analyzed to verify the results. RESULTS: TIM-3 expression of CD34(+) leukemic blasts (R(2 )= 0.95, p<0.0001) and CD34(+)CD38(-) leukemic stem cells (R(2 )= 0.75, p<0.0001) were significantly and positively correlated with that of the whole population of leukemic blasts. In addition, TIM-3 expression level of leukemic blasts correlated significantly and positively with that of CD8(+) (R(2 )= 0.44, p<0.0001) and CD4(+) (R(2 )= 0.16, p=0.0181) lymphocytes, and higher TIM-3 expression of leukemic blasts was significantly associated with a greater proportion of peripheral CD8(+) T lymphocytes (R(2 )= 0.24, p=0.0092), indicating that TIM-3 on leukemic blasts might alter adaptive immunity of AML patients. Regarding clinical data, the presence of core binding factor (CBF) translocations was significantly correlated with higher TIM-3 expression of leukemic blasts (CBF versus non-CBF, median 22.78% versus 1.28%, p=0.0012), while TIM-3 expression levels of leukemic blasts were not significantly associated with the remission status after induction chemotherapy (p=0.9799), overall survival (p=0.4201) or event-free survival (p=0.9873). Similar to our results, TCGA data showed that patients with CBF translocations had significantly higher mRNA expression level of HAVCR2 (the gene encoding TIM-3) (median, 9.81 versus 8.69, p<0.0001), and as all patients in the cohort were divided into two groups based on the median HAVCR2 expression level, 5-year overall survivals were not significantly different (low versus high, 24.95% versus 24.54%, p=0.6660). CONCLUSION: TIM-3 expression level on AML blasts correlates with presence of CBF translocations rather than clinical outcomes.