Wonder Drug Strikes Again!

Abstract & Commentary

By Jonathan Abrams, MD Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque. Dr. Abrams serves on the speaker’s bureau for Merck, Pfizer, and Parke-Davis.

Synopsis: Use of statins in patients with atherosclerosis is associated with a reduced risk of subsequent sepsis. Randomized trials of statins for prevention of sepsis are warranted.

Source: Hackam DG, et al. Statins and Sepsis in Patients with Cardiovascular Disease: A Population-Based Cohort Analysis. Lancet. 2006;367:413-418.

Several observational studies, as well as some animal research have suggested a beneficial role of statins in subjects with acute bacterial infection and sepsis. The hypothesis that statins may manifest non-lipid actions (pleiotropic effects) has been a popular theme for years. These drugs have effects on immune regulation, inflammation, and thrombogenesis, in addition to anti-oxidant actions and augmentation of nitric oxide. This Canadian study from the province of Ontario, postulated that statins may reduce the incidence of sepsis in a population of subjects with atherosclerosis.1 This trial was a large scale, 5 year, population-based cohort study, and comprehensive analysis of sepsis events using propensity-matching to minimize confounding. A sophisticated statistical and epidemiological base underscores this study. Hackam and colleagues from Sunnybrook and Women’s College Health Sciences Center in Toronto created a retrospective patient cohort of elderly individuals (> 65 years). They used 4 large validated data bases to identify appropriate individuals, and tracked consecutive patients admitted for a cardiovascular event or who had a revascularization procedure at any hospital in the province of Ontario between 1997 and 2001. All patients had acute coronary syndrome, acute ischemic stroke, or carotid or coronary revascularization procedure. Surveillance for statin use was carefully carried out by the Ontario Drug Benefit; all elderly patients in Ontario received universal prescription drug coverage through this formulary. Statin users were defined as individuals who had one of more prescriptions written during the 3 months preceding the index date, heralding the start of the analysis; nonusers had no statin prescription during this period. Propensity-based matching was utilized to establish a control subject for each individual who received a statin. The observation period began 3 months after hospitalization for each patient, and continued until death, hospital admission for sepsis, or termination of the study (March 31, 2002). The follow-up interval was subsequently extended to March 31, 2004. All sepsis admissions were identified. The primary analysis evaluated patients receiving a statin vs no statin. Eight pre-specified subgroups were identified who were felt to be at higher risk for development of sepsis. The association of statin use with the composite outcome of death, acute MI, and ischemic stroke was evaluated to, "check for the presence of an association where one would be expected."

Results: During the 5 year study, 173,000 individuals were hospitalized for vascular events. There were approximately 141,000 survivors at 90 days after discharge, one third of whom had received a statin during this period. Propensity-based matching was then carried out to create a final cohort of approximately 70,000 patients, half of whom received an initial statin prescription and half who did not. Statin users and controls were very similar withrespect to demographic characteristics, sepsis risk factors, and other healthcare markers. Diabetes, congestive heart failure, and COPD were common. Mean follow-up was 2.2 years, during which time 550 patients were admitted for sepsis in the statin group and 667 in the control group. Thus, the rate of sepsis was significantly lower for statin treated patients (71 vs 88 per 10,000 person/years, P = 0.003). The crude hazard ratio was 0.81, which was statistically significant. After adjustment for a variety of sepsis risk factor co-morbidities, there was still a 19% reduction in sepsis events. When adherence to therapy was evaluated, the association between statin use and sepsis was accentuated. Statin patients had fewer episodes of sepsis during the extended analysis and mean follow-up of 3.8 years. They also had a lower risk of severe and fatal sepsis than did controls. There was no obvious dose relationship with respect to the protective effects. Statin use was associated with a decrease in the composite of death, MI, and stroke (12% reduction). Non-statin lipid lowering drugs had no effect on sepsis incidence. Hackam et al conclude that the, "apparent protective association between statins and sepsis was consistent across several high-risk groups, was apparent throughout the entire follow-up period, and was amplified in analyses accounting for non-adherence and crossovers."

Hackam et al discuss statins and sepsis data and point out that their results are concordant with human and animal data on statins and sepsis; in 2 published studies of fewer than 400 patients each, statins had a marked salutary effect on sepsis mortality and progression of infection due to severe sepsis. In an animal model of sepsis, statin treatment led to "substantial reductions in inflammatory cytokines and activation of immune cells." This has also been found in humans. Statins have been shown to reduce excessive nitric oxide associated with shock and circulatory collapse, and appear to help maintain favorable hemodynamics in an animal model of polymicrobial sepsis. Furthermore, data has been positive for statin effects on several other bacterial, fungal, and viral pathogens. Hackam et al point out that their results are probably applicable to individuals without symptomatic atherosclerosis, such that cardiovascular co-morbidity is unlikely to be an important interaction. They conclude that statin use "was associated with a significantly reduced risk of sepsis, including severe sepsis and fatal sepsis." They suggest that individuals who are taking statins not stop their medications during serious infections, and that statins should not be discontinued routinely before high risk surgery. They even suggest that, "statins might be considered for patients at very high risk for sepsis," particularly if there is underlying cardiovascular disease. And finally, they call for future randomized control trials.

An accompanying editorial commentary is supportive of the study findings that there may be potential anti-inflammatory actions that are independent from the lipid effects of statins and that might be important in a variety of infections.2 Hackam et al suggest that statin efficacy in sepsis is probably independent of lipid lowering actions, and is likely "to be mediated by interference with isoprenoid synthesis and subsequent geranylation of membrane proteins." Other studies have suggested anti-chlamydial action, as well as possible effectiveness against fungal and viral pathogens. They also call for large, placebo-controlled, randomized trials to assess statin prevention of sepsis, as well as treatment of overt sepsis. They emphasize the potential beneficial effect of these drugs in other arenas, including viral infections and even neurodegenerative disorders.2


The hypothesis that statin drugs or HMG Co-A reductase inhibitors may have important biological effects in addition to, or separate from, LDL cholesterol lowering, has been around for years. There is an increasing number of manuscripts published suggesting statin efficacy in a variety of unrelated conditions, including congestive heart failure, atrial fibrillation, aortic stenosis, dementia, and the treatment of sepsis. If one examines the biochemical pathway beyond the HMG Co-A reductase step, it is clear that a variety of other processes may be affected when the parent enzyme is blocked. For instance, a rho-kinase inhibitor is being evaluated as a treatment of angina; this is an example of an interaction with moieties distal to the HMG Co-A reductase interaction. Prenylation of various proteins are affected by the use of statins, and this may have far reaching effects. A number of review articles have been published on this subject; 3 recent commentaries are recommended to the interested reader.3-5

In this careful, ambitious, and very large epidemiologic analysis of statin utilization in Ontario province, Hackam et al wanted to see if exposure to a statin would have a salutary effect of serious bacterial ( and perhaps other agents) infections and, in particular, sepsis. Using an unusual statistical design, they established a control group and a statin group of 35,000 patients each, and extracted information for an average of 2.2 years of subsequent evaluation for the development of sepsis and septic shock. Statin use did not have to be sustained, but was based on whether a statin was included in a patient’s regimen several months after index hospitalization for an atherosclerotic event. The population studied had an acute coronary or cerebral event, or underwent revascularization, thus providing an appropriate rationale for statin utilization. Although only one third of the population was exposed to statins during the study period, the 2 equal groups, selected with the aid of a propensity analysis, appear to be large enough to produce a sufficient number of events for meaningful analysis. The degree of exposure and the dose and potency of statins are not available. Furthermore, the precise mechanism(s) of benefit cannot be ascertained from this study. The finding of an approximate 20% reduction in serious bacterial infections and sepsis in statin treated patients is remarkable. The type and nature of organisms are unknown. One must assume that sepsis was appropriately defined in the hospitals. Hackam et al,1 as well as Merx and colleagues in an editorial commentary,2 call for a randomized trial to further test the hypothesis that an HMG Co-A reductase inhibitor could have a beneficial effect on serious infections. The data certainly suggest this, although it does not seem appropriate to prescribe a statin as therapy in patients with sepsis until this hypothesis is accepted by the general medical community. Nevertheless, this is an exciting development, outpacing all of the other current hypotheses for pleiotropic statin effects in non-cardiac conditions. Hopefully, clinical trialists will quickly undertake an appropriately sized, placebo-controlled study that will resolve this question.


1. Hackam DG, et al. Statins and Sepsis in Patients with Cardiovascular Disease: A Population-Based Cohort Analysis. Lancet. 2006;367:413-418.

2. Merx MW, Weber C. Statins: A Preventive Strike Against Sepsis in Patients with Cardiovascular Disease? Lancet. 2006;367:372-373.

3. Robinson JG, et al. Pleiotropic Effects of Statins: Benefit Beyond Cholesterol Reduction? A Meta-Regression Analysis. J Am Coll Cardiol. 2005;46:1855-1862.

4. Ray KK, Cannon CP. Early Time to Benefit with Intensive Statin Treatment: Could It Be the Pleiotropic Effects? Am J Cardiol. 2005;96:54F-60F.

5. Ray KK, Cannon CP. The Potential Relevance of the Multiple Lipid-Independent (Pleiotropic) Effects of Statins in the Management of Acute Coronary Syndromes. J Am Coll Cardiol. 2005;46:1425-1433.