ALBERT

Description: Background: Continuing growth in the incidence of leukemia suggests a possible environmental etiology correlated to the increase of environmental pollution. Recently, environmental particulate pollution (EPP) has been declared by IARC a Class I carcinogenic agent; it looks reasonable to presume that not only chemicals like benzene and its derivatives, but also other components like EPP could be worth of study. No specific researches have up to now focused the role of EPP on acute myeloid leukemia; we thus have identified a suitable instrumentation and protocol to show the presence and composition of particulate matter in blood samples of patients affected by acute myeloid leukemia patients and in healthy controls. Methods: 38 peripheral blood samples (19 acute myeloid leukemia, 19 healthy controls) were analyzed by means of an Environmental Scanning Electron Microscopy (ESEM) coupled with an Energy Dispersive Spectroscopy (EDS) a sensor capable of identifying the composition of micro- and nano-particles of exogenous nature in pathological tissues (applied for the first time in the current study on blood samples). The results were statistically treated with unpaired two-tailed Student's t-test, MANOVA and Principal Component Analysis. Results: A consistent quantity of micron-, submicron- and nano-sized foreign bodies (from 20 micron down to 100nm) was documented in 18/19 AML cases, whereas they were absent or rare in the controls. The particles appeared as singlet and aggregates (ranging from 5 to 20micron), either in close contact with blood elements or interacting with plasma. Some reacted with blood proteins thus forming composite clusters. A total of 141 aggregates (median 8, range 0-18) in AML, compared to a total of 12 aggregates in controls (median 1, range 0-3) were counted. The aggregate analysis showed variable sizes and number of particles, with a total of 5394 particles in leukemia cases compared to a total of 207 in controls. The total numbers of aggregates and particles were statistically different between cases and controls (MANOVA, P