Pharmaceutical agents aimed at reducing tumor necrosis factor-alpha (TNF-α) levels appeared to be attractive possibilities in the medical management of heart failure, as heart failure was shown to be associated with high TNF-α levels. However, therapies specifically targeting TNF-α failed to show any survival benefit. We examined whether a broad inhibition of inflammatory cytokine production secondary to macrophage inhibition would be more effective at improving cardiac function in the setting of heart failure. To this end, we studied Semapimod (formerly known as CNI-1493), a synthetic guanylhydrazone that inhibits macrophage activation and the production of several inflammatory cytokines. Left anterior descending coronary ligation surgery was performed on each animal to induce a myocardial infarction. After confirming heart failure by echocardiography, the animals were randomly assigned to one of four groups: (1) rats with myocardial infarct receiving high-dose Semapimod, 10 mg/kg/d (MI-H, N= 13); (2) rats with myocardial infarct receiving low-dose Semapimod, 3 mg/kg/d (MI-L, N = 9); (3) rats with myocardial infarct receiving vehicle treatment, 2.5% mannitol in water (MI-0, N = 9); and (4) control rats with sham operation and vehicle treatment (Sham, N = 10). Both Semapimod and vehicle treatments were administered by daily tail vein injections over a course of five days. Echoes were repeated at 2, 5, and 9 weeks following treatment. At 9 weeks, hemodynamic data were collected and the animals were euthanized. Trichrome staining was done to assess infarct, and immunohistochemistry was performed to assess TNF-α levels. Prior to drug administration, serum was taken from 5 random rats. No detectable level of TNF-α was seen (lower limit of detection for the assay used = 12.5 pg/mL). Also prior to any treatment, echocardiography confirmed significant cardiac impairment of rats undergoing LAD ligation as compared with sham. Over the course of the 9 weeks, there were 4 deaths, all in the MI-H group. There was no difference between Semapimod-treated animals and vehicle-treated MI animals in any echocardiographic or hemodynamic parameter. TNF-α staining in the noninfarcted region was evident only in the MI groups, not the sham group. When blindly compared on a semi-quantitative scale (ie, 0 = no visible staining to 3 = marked staining), a significant difference in staining was observed between MI-0 versus MI-H (1.19 ± 0.32 versus 0.33 ± 0.14; P = 0.03) and between MI-0 and MI-L (1.19 ± 0.32 versus 0.39 ± 0.22; P = 0.05). In this setting, despite the fact that Semapimod treatment decreased tissue TNF-α levels, it did not improve cardiac function, and at high doses it was associated with higher mortality. These results in a rodent model confirm the results of clinical trials with etanercept and infliximab (ie, that decreasing TNF levels in plasma or tissues does not improve cardiac function and may actually increase mortality).