Ten years apart, almost to the day, the results of two clinical trials of "magic bullets" for the treatment of sepsis were published in the New England Journal of Medicine.1-2 Both studies claimed to reduce relative mortality by 19 percent. Each study was reported to be the first trial that would allow for the use of a medicine (a.k.a. magic bullet) in a disease that claims the lives of over 750,000 patients per year in the United States alone. Both pharmaceutical companies hired large sales forces and provided these agents for compassionate use prior to FDA approval. An impression of déjà vu occurred following rumors of anticipated drug costs of "several thousand dollars per course of therapy" that could double hospital pharmacy budgets. As it did ten years ago, the potential financial impact of these life-saving drugs also created concern for pharmacy directors, provided hot topics for Pharmacy and Therapeutic Committees and caused potential resource problems for Chief Fiscal Officers of hospitals. A priori use guidelines became a priority in anticipation of pending approval of both agents.
The human monoclonal antibody to endotoxin (HA-1A: Centocor) provided an apparent benefit to patients with gram negative infections in shock.1 The 1991 N Eng J Med article documenting the results of the clinical trial with HA-1A created a flurry of activity after the drug was released in Europe at a price of $3,750 per dose. Once experts in the literature, the FDA and virtually every P & T committee, ICU Director and infectious disease practitioner in the country scrutinized the study, the drug was driven into extinction.3-6 No one could repeat the endotoxin binding studies, questions were raised about apparent deficiencies of the antibiotic treatment in the placebo arm, and rumored flaws in the conduct of the protocol and a failed demonstration in a repeat clinical trial all contributed to lost hope for the use of HA-1A.7
The rest of the 1990's were spent searching in different directions for the elusive magic bullet. A seemingly rational approach that theorized blocking the over-exaggerated host immune pro-inflammatory response would benefit patients.8 The hype of this theory led to the development and clinical testing of a variety of agents in over 6200 patients. A meta-analysis of these trials showed a small beneficial treatment effect; however, the major effect of more than 30 trials has been the mortality of a number of the biotechnology companies.9
Subsequent trials evaluating therapies that bound endotoxin or attempted to block the excessive innate immune response to sepsis have failed. The major explanation for this unfortunate lack of success may be attributed to "locking the barn door after the horse has been stolen." In other words, human clinical trials with anti-cytokine, anti-endotoxin, cytokine receptors or antagonists, or anti-inflammatory therapies all have the fundamental flaw of treating well after the targets have been released and triggered a very complex series of events and cascades. The problem of patient heterogeneity, finding subjects with similar disease characteristics, also has complicated positive results despite promising animal and small phase II human trials.
Much interest has developed in the last few years investigating the relationship between inflammation and coagulation.10-13 The Systemic Inflammatory Response (SIRS) to infection leads to the activation of a variety of host innate immune responses. This SIRS response generally involves the release of a variety of pro- and anti-inflammatory cytokines, proteolytic enzymes and other mediators that mount a systemic reaction to the offending pathogen and virulent by-products. As this process develops, tissue factor is released from macrophages and monocytes, which in turn activates the extrinsic coagulation cascade. This process then up-regulates itself and full activation of the coagulation system is achieved with the involvement of a variety of components including platelets, thrombin, neutrophils and the endothelium itself. The activation of various intrinsic factors causes the generation of thrombin from prothrombin, which catalyzes the conversion of fibrinogen to fibrin. This fibrin can then accumulate and can be deposited in the microvasculature, leading to vascular injury, organ ischemia and dysfunction multi-organ failure and death.
There is increased enthusiasm for the concept that the various coagulation proteins have anti-inflammatory effects as well. Excess thrombin becomes available when the natural serine protease cascade elements become depleted. The excess thrombin, tissue factor, platelet activating factor, thrombomodulin and other factors appear to up-regulate the pro-inflammatory response as measured by Il-6 and D-dimers. A variety of clinical trials have and are being conducted to evaluate this theory.
Three different endogenous inhibitors of coagulation, antithrombin (AT), tissue factor pathway inhibitor (TFPI) and recombinant activated protein C (APC) have been extensively studied.
In a series of small studies, antithrombin (formerly antithrombin III: actually no one could find ATI or ATII), has been the most widely studied inhibitor of coagulation. It is theorized to work by binding excess thrombin and has the advantage of measurable serum activity. The majority of work has been done with pooled plasma products. A recent trial in Europe using a pooled plasma-based antithrombin product vs. placebo did not show a mortality benefit in sepsis.14 The trial did not control the use of heparin, which may have affected its efficacy. The issue of proper dosing or appropriate replacement of endogenous antithrombin activity appears to be an issue as well.
Tissue Factor Pathway inhibitor (TFPI, tifacogin: Chiron) also has shown promise in in-vitro, in-vivo animal and human trials. It directly binds tissue factor, which may be a more specific method of reducing or controlling the inflammatory response.15-16 TFPI has recently concluded testing in a large scale Phase III clinical trial; 17 patients were studied at MGH. The results should be available soon.
A third endogenous inhibitor of coagulation is recombinant Human activated Protein C (rhAPC, Drotrecogin alpha: Xigris). In sepsis and other inflammatory conditions, Protein C is activated.17 Activated protein C binds to thrombomodulin, which in turn binds thrombin. Deficiency of activated Protein C leads to increased thrombin availability, an increased inflammatory response and increased mortality. The PROWESS study replaced Activated protein C, hoping to restore hemostasis and reduce mortality from severe sepsis. The result of this trial of a possible "magic bullet" has created strikingly similar attention in 2001 to HA-1A a decade ago.
Enthusiasm and hope has been generated based on the results of this large Phase III clinical trial in sepsis.2 Investigators from 164 centers in 11 countries gave rh-APC or placebo to patients in severe sepsis for 96 hours. In the PROWESS study there was a significant reduction vs. placebo in all-cause mortality in a severe septic population. The trial showed a 6.1% absolute reduction in mortality with a relative reduction of 19.1%.
A variety of issues must be addressed in identifying appropriate patients to treat.
The population that benefited from rhAPC therapy was fairly ill as evidenced by: APACHE II baseline of 25, of which 75% had 2 organ failures. However, 80% of patients were from the community 87.6% had depleted Protein C at baseline and 53% had pneumonia indicating a select or homogenous group of patients, particularly those in shock, showing benefit. These represent possible targets for optimal clinical use. There are possible arguments against using this drug for "nosocomial sepsis", or patients who had been in the hospital and already had organ failure from previous organ dysfunction when a new septic event occurs. Less ill septic patients (stratified by APACHE II < 19), patients with thrombocytopenia or severe liver disease, and patients with poor prognosis from underlying pre-septic conditions must be carefully screened and probably excluded from receiving rhAPC.
All patients were carefully screened for any high risk for bleeding. Despite the bleeding risk screen, there was still an increased bleeding risk in the patients who received therapy with drotrecogin alpha. These bleeds, however, were not considered serious except in 2 patients in the rhAPC group and 1 in the placebo group (intra-cranial bleeds). Specific patient populations requiring therapeutic anti-coagulation and those needing anti-thrombotic therapy or thrombolysis or are at an increased risk for bleeding will need to be carefully identified and excluded now that drotrecogin has reached clinical use. In a subsequent open label trial, the intracranial bleeding risk in patients with thrombocytopenia has increased to 2.1%. Transcripts and information from the 10/16/2001 FDA advisory board meeting with additional safety and efficacy information is available at http://www.fda.gov/ohrms/dockets/ac/cder01.htm#Anti-Infective.
The financial resource impact is an important issue which must also be addressed. One method of projecting the impact of this trial is to evaluate its benefit in relation to number of patients who may qualify in the "real" clinical arena. The number of patients needed to treat with rhAPC to save 1 life is 16.4. If you multiply the estimated cost of greater than $5000 per course of therapy, this equals at least $80,000 to save one life. This can be compared to the GUSTO trial, which showed the number of patients needed to treat of 100 with tissue plasminogen activator over streptokinase.18 This translates into greater than $180,000 to achieve a 1% reduction in mortality. Tissue plasminogen activator therapy has become the standard of care in acute myocardial infarction, with a substantially less mortality benefit than rhAPC.
The study appears to have been well designed and executed and the drug has recently been approved. This trial offers ammunition for the argument that replacing natural coagulation factors has a role in preventing the organ dysfunction and mortality in septic patients. The challenge for Pharmacy and Therapeutic Committees, Critical Care and Infectious Disease practitioners will be to extrapolate inclusion/exclusion criteria and results from one clinical trial into fiscally responsible guidelines for appropriate use.
These exciting recent results appear to offer hope for our patients who develop severe sepsis. The process of getting these magic bullets into the hands of health care providers will certainly include some hype. Hopefully, as guidelines are developed that combine rationale scientific knowledge, fiscal responsibility and multi-disciplinary collaboration, hysteria will be avoided.
Portions of this review have been previously published in the Society of Infectious Diseases Pharmacists Newsletter: Spring 2001 edition (Available on-line at www.sidp.org)
Selected References: