To evaluate the therapeutic effects of the HELP system (heparin-induced extracorporeal low-density lipoprotein-apolipoprotein(a)-fibrinogen precipitation) in patients with acute ischemic stroke and probe into its possible mechanism, 10 patients with acute ischemic stroke were included in this study and received the HELP therapy in addition to low molecular weight dextran, salvia miltiorrhiza and aspirin. Matched with gender, age, European Stroke Scale (ESS) and fibrinogen, 20 patients with cerebral infarction treated with low molecular heparin were chosen as controls during the same period. We compared the efficacy and prognosis between the two groups. In order to clarify the effects of HELP treatment on endothelial function, human umbilical vein endothelial cells (HUVEC) were cultured with the serum of pre- and post-HELP therapy and control patients, then endothelial permeability and biomarkers of endothelial dysfunction or activation in the supernatant of incubated HUVEC, such as soluble thrombomodulin (sTM), soluble intracellular adhesion molecule-1 (sICAM-1) and monocyte chemoattractant protein-1 (MCP-1), were tested. Twenty-one days after treatment, ESS scores in the HELP group (70.4 +/- 23.06) were higher than those in control group (60.7 +/- 18.94), but there was no statistical significance (P > 0.05). In the HELP group, total efficacy rate reached 60% with ESS score (6 cases of recovery and efficacy, 4 cases of inefficacy), while it was 40% in control group (8 cases of recovery and efficacy, 12 cases of inefficacy), but unfortunately there was no significant difference between both groups using the chi(2)-test (P > 0.05). Interestingly, with the activities of daily living (ADL) score, the efficacy rate in the HELP group (60%) was markedly higher than that in the control group (20%) (P < 0.05). Furthermore, Pearson correlation analysis showed that the therapeutic window (the time from the patient's onset to receiving the therapy) was correlated to final ADL scores (P = 0.044) and the mean therapeutic window was 14.08 +/- 3.41 h in patients who achieved efficacy criteria. After therapy, no difference was found in hemorheology, fibrinogen, blood lipid, oxidized low-density lipoprotein (oxLDL) or C-reactive protein (CRP) in the control group, while there was a significant decrease in the HELP group immediately after treatment (P < 0.01). But the hospital stay time was similar between the HELP group and the controls (23.11 +/- 10.65 vs. 21.53 +/- 8.73 days; P > 0.05) and 21 days later, the dimension of the largest cross-sectioned infarction area by CT scan did not significantly change for the two groups' of patients (P > 0.05). When cultured with the patient's serum after HELP therapy, the concentration of sTM, sICAM-1, and MCP-1 in the supernatant of cultured HUVEC remained unchanged at 24 h (P > 0.05), while it was remarkably higher when HUVEC was cultured with the serum of the controls and patients before HELP therapy (P < 0.01), except for sTM. The endothelial permeability was also improved after the HELP treatment. With the effects of improving hemorheology, decreasing acute phase reactive proteins such as CRP and fibrinogen, ameliorating endothelial cell function, HELP system may be a novel therapy for acute ischemic stroke.