Early detection for protection

Early identification of influenza and hantavirus

One of the new projects is based on the detection of signature proteins that decorate the surface of the influenza virus particle. In collaboration with University of New Mexico Medical School, Lab researchers are developing a compact sensor system using thin films that mimic cell membranes and wave-guide-based optical detection. This small hand-held device will be simple to use and capable of detecting influenza early after infection. The project's objective is a simple inexpensive device that could be used in a doctor's office to guide treatment or in the field to provide information on the spread of the infection.

The initial effort for this detector is focused on influenza and hantavirus. Influenza is one of the leading causes of death in the United States, and health experts are predicting a likely occurrence of a virulent new influenza strain striking in the near future, similar to the 1918 pandemic that killed more than 20 million people worldwide. Hantavirus, which is common in many South American countries, has now been found in more than 25 states, and its hemorrhagic fever kills approximately 45 percent of its victims. In many cases, at rural clinics patients with flu-like symptoms are sent home with a diagnosis of influenza, only to return in shock and quickly succumb to the infection before proper treatment can be initiated.

With early diagnosis, patient survival rates improve significantly because they quickly can be referred to advanced care facilities for appropriate treatment.

Infection from many biothreat agents (e.g., Bacillus anthracis that causes anthrax) also produce flu-like symptoms. Accordingly, in the event of a bioterrorist attack, one of the problems facing first responders and health-care workers is to be able to distinguish bioagent infection from influenza. Without the disease's early identification, the patient would likely be sent home only to die days later when the infection moved beyond the treatable stage. A robust hand-held device that could quickly screen potential victims to distinguish between influenza and infection or exposure to a biothreat agent is critical for triage and will help save lives.

Early Warning — A prognostic epidemiology network

The need for rapid identification combined with the ability to swiftly notify other members of the health-care community forms the basis of a second Los Alamos effort, that of an integrated early warning system for influenza. The researchers are working to combine the skills of chemistry, biology, space science and systems analysis to anticipate the course of an influenza outbreak and speedily mitigate or eliminate its impact. Talent in genomics, cell biology, sensor development, modeling and simulation, statistics and advanced systems engineering will all come into play for this project, as will relationships with the Centers for Disease Control and Prevention, the UNM School of Medicine and the New Mexico Department of Health. Once perfected in the influenza application, the components will then be adapted to a range of other natural and terrorist-derived threats.

Such a program requires a multi-level approach, an approach that Los Alamos is an acknowledged leader. Among the components of the system will be the following:

• new data sources for influenza strain identification,

• new computer models for predicting disease spread and evolution and

• a nationwide network for communicating key informa- tion and planning a response.

The first part of this effort, early identification of influenza strains, will leverage the work already described, namely the development of a small hand-held device capable of detecting influenza early after infection. Further develop- ment will identify genetic markers that are associated with virulence of the virus and incorporate those into models of influenza evolution and transmission. This research can provide guidance as to which measures are most effective in containing an impending outbreak.

In part two, detection sensor research will be expanded to identify and detect toxins, viruses and DNA strands at vanishingly small concentrations. The sensor described previously uses a fluorescence-based recognition system. Another promising effort already afoot, tags nucleotide ecognition elements to allow identification of viral RNA sequences. Yet a third could be effective as well, that of using "molecular beacons" that each detect a different target, marking single nucleotide variations in different virus strains.

Part three goes beyond the challenge of seeing what an infected patient is battling. In this phase, the objective is to determine what the patient has been exposed to, even before symptoms are evident. Chemical changes within the epithelial lining of the lungs appear to provide useful information in this area. If patients can be identified before they are either sick or infectious, simple public-health steps such as isolating and breaking the expanding cycle of infection before it begins possibly can stop an impending epidemic.

The final phase of this aggressive program is called a Multi-Level Heterogenous Data Fusion. It addresses the crippling isolation of the very health-care facilities that find and treat the earliest victims of an outbreak or biological attack. Early detection of influenza-like illnesses is critical to the nation's ability to detect and respond to a biological attack. Yet current medical reporting techniques involve a health-care worker writing a case analysis some hours or days after a patient is seen. While the potential exists to gather extensive data from the scattered health-care facilities, until these facilities are networked and reporting information in a more timely, structured fashion, information is too little, too late.

On the local scale, health-care providers immediately need to recognize unusual events and take appropriate action. Regionally or nationally, the local information must be transmitted to the decision-makers quickly enough that mitigating measures can be put into place.

Existing prototype systems at Los Alamos can be tested to determine if they are of use in filling these information gaps. Current Los Alamos projects such as the Biological Aerosol Sentry Information System and a medical surveillance system, called Rapid Syndrome Validation Project, (see "Rapid Syndrome Validation Project ") both demonstrate the Laboratory's experience with networked sensor and information systems and the ability to work with industry to provide open standards for combining clinic information systems. Research that uses tools such as Mobile Software Agents that can perform detailed pattern searching on data that educes an impenetrable mass of information to the key items can tell users whether or not a virulent outbreak of the flu is at hand.