The scientific concept of what sepsis is has changed over the decades. But unfortunately, to date none of the changes have made a dent in its mortality rate, which continues to hover at about 30 percent.

"Sepsis was originally thought to be an infectious disease problem," William Aird told BioWorld Today. Logically, the best treatment under that theory was vigorous antibiotics.

"Then it became apparent that no matter how well you treat the bug and how well you support the patient, you are still left with that 30 percent mortality rate," said Aird, who is the chief of the division of molecular medicine at Beth Israel Deaconess Medical Center.

It slowly was recognized that sepsis is the host's immune response to infection, gone out of control - "a great weapon that turns on itself."

While the basic insight was correct, the treatments that followed were not. Aird dryly noted that "for the longest time, there was very little evidence-based medicine in intensive care." Just about any molecule whose levels are elevated in septic patients has been the target of a clinical trial at some point. The result? Activated protein C to treat sepsis, plus dozens of failed clinical trials.

Activated protein C "is sort of like a cluster bomb," Aird said, attacking multiple steps in the pathway. And while it's hard to argue with success, especially with the only success there is to date, the current thinking is evolving yet again. Now sepsis is thought of as a network of responses, and the best treatment is one that will take out the hubs of that network.

"Taking out the O'Hare's or the JFK's - if you can target them, that's the best hope for single-modality treatment," Aird said.

Aird and his colleagues think they have found one such hub: vascular endothelial growth factor, or VEGF.

In the June 12, 2006, issue of the Journal of Experimental Medicine, scientists from Beth Israel; McMaster University in Hamilton, Ontario; and the University of Leuven in Belgium reported on experiments that suggested VEGF could cause, rather than correlate with, the outlandish immune response that is sepsis.

In several mouse models, inducing sepsis led to elevated levels of VEGF. Preventing VEGF from binding to its receptors reversed sepsis symptoms from elevated cytokines to organ damage, and lowered mortality from more than 50 percent in untreated controls to about 25 percent.

Animals that overexpressed VEGF had 100 percent mortality after being injected with bacterial endotoxin; in contrast, animals overexpressing a soluble binding protein that prevented VEGF from interacting with its receptors were somewhat protected against sepsis, with a mortality rate of about "only" 25 percent.

Through antibody studies, the scientists identified the specific receptor most likely to be mediating VEGF's deadly effects. The relationship between endotoxin and VEGF was something of a vicious cycle. Endotoxin elevated VEGF levels, but animals overexpressing VEGF also were especially sensitive to endotoxin.

Aird and his colleagues also used two human models - sepsis patients and healthy volunteers. The researchers measured the levels of VEGF and the related placental growth factor in sepsis patients and noted that both were elevated. They also injected volunteers with bacterial endotoxin to produce sepsis-like symptoms; those injections led to elevated levels of both growth factors.

From a purely experimental perspective, the human model obviously stops far short of true sepsis. But Aird, who noted that every model has its advantages and shortcomings, said that "the injection of bacterial product - not bacteria themselves - elicits an innate immune response that is comparable, though not identical, to what we see in our patients with severe sepsis," adding that the human model is "as close as you can get to an ethically sound induction of an immune response."

VEGF seems like it should be easy to partner - South San Francisco-based Genentech Inc.'s Avastin certainly has shown that it can be a clinically and financially profitable enterprise, and the latest new company focused on VEGF technology was announced just last month. (See BioWorld Today, May 9, 2006.)

But Aird said that he and his colleagues still are looking for industrial partners; they are in discussion but have yet to close a deal. VEGF expertise by and large goes hand in hand with a focus on cancer.

And there is, of course, the matter of the 30 or so failed Phase III trials. TNF-alpha, interleukin-1 - "all worked in animals and failed in humans," Aird acknowledged. Nevertheless, given his team's experimental results, he described himself as "anxious" to target VEGF in humans.

"We really believe this could be the thing that works," he said.