Antimicrobial Resistance Diagnostic (AMR) Challenge Step 2 Finalists

Principal Investigator: Richard Anderson
Institution: Becton, Dickinson and Company, Franklin Lakes, New Jersey
Executive Summary: Becton, Dickinson and Company (BD) seeks to reduce the unnecessary use of antibiotics by guiding healthcare practitioners in differentiating between bacterial and viral infections within the time constraints of a single outpatient visit. Building on the public-health utility and broad commercial success of the BD Veritor™ platform, we are developing a novel digital immunoassay (DIA) that will measure a panel of circulating host immune biomarkers from a single blood-based sample. The test is intended to be an aid in the management of patients with symptoms of infection – such as upper respiratory infection – for which antibiotic misuse is a widespread problem. By providing objective diagnostic evidence that the patient’s immune system is fighting either a bacterial or a viral infection, the test will return an informative result without requiring a sample from the site of infection, and without the time or expense of diagnostic confirmation of numerous possible candidate pathogens. By employing the sample-in/result-out, one-button simplicity of the CLIA-waived BD Veritor system, the test can be deployed broadly at point-of-care settings without the need for trained personnel. And by returning a result rapidly (target of ten minutes) at the point of care, the test outcome will enable an actionable decision before the patient leaves the site.

Principal Investigator: Sophia Koo
Institution: Brigham and Women's Hospital, Boston, Massachusetts
Executive Summary: Community acquired pneumonia (CAP) is the fourth leading cause of death worldwide, with 430----450 million cases and 3----4 million deaths per year. In the United States alone, there are 4----5 million CAP cases, 1.1 million hospitalizations, and over 52,000 deaths annually, accounting for over $34 billion in direct health expenditures. Despite the high incidence of CAP, existing diagnostic tests for CAP are inadequate, identifying an underlying microbial pathogen in only 7----38% of cases. Even when a specific causal pathogen is identified, results often return long after it is necessary to make a clinical treatment decision. Because of the difficulty of identifying the microbial cause of pneumonia and the potentially severe consequences of not treating bacterial pneumonia promptly, clinicians often treat CAP empirically with antibiotics, even though a large proportion of these cases are viral in etiology. Unnecessary broad----spectrum antibiotic prescriptions per year, in turn, fuel the emergence and spread of resistant bacteria, including most of the urgent and serious threat level pathogens on the Center for Disease Control and Prevention’s list of urgent and antibiotic---resistant threats, on a population level. We propose a novel approach to CAP diagnosis based on distinct differences in the breath volatile metabolite profiles of patients with bacterial vs. viral pneumonia, using a rapid, portable gas chromatography--- differential mobility spectrometry (GC----DMS) Microanalyzer device at the point of care with parallel gold standard thermal desorption gas chromatography----tandem mass spectrometry (GC----MS/MS) in the laboratory to detect specific volatile metabolite signatures that (a) differentiate bacterial vs. viral CAP, and (b) identify certain common, specific causes of CAP, including Streptococcus pneumoniae. This rapid breath test for CAP would guide antimicrobial treatment decisions at the point of care, reducing antibiotic prescribing in patients who do not have bacterial pneumonia and slowing emergence of antimicrobial resistance on a population level.

Principal Investigator: Don Straus
Institution: First Light Biosciences, Inc., Bedford, Massachusetts             
Executive Summary: First Light Biosciences is addressing two critical issues in healthcare: the rise in superbugs that are re-sistant to most antibiotics and the epidemic of hospital acquired infections. Today, a large fraction of life threatening infections seen in hospitals are caused by pathogens that are resistant to multiple anti-biotics. Effective treatment and cure requires rapid diagnostics to detect infections and select appropri-ate antibiotics for treatment. Unfortunately, no such tests exist today. We propose to develop tests based on First Light’s MultiPath technology that can rapidly and affordably detect a broad range of in-fections, accurately identify the infectious agents, and determine the appropriate antibiotic for treat-ment. The MultiPath platform consists of an automated benchtop analyzer that accommodates a broad menu of application-specific consumable cartridges. The MultiPath technology is quantitative. In only 15 min it detects infections and identifies pathogens including bacteria, fungi, viruses, toxins, human cells, and biomarkers. The technology can determine the correct antibiotic therapy in just 4 hours in contrast to the several days required by current methods. While as sensitive as high performance cen-tral laboratory tests, the MultiPath tests are as rapid and as easy-to-use as point-of-care tests and re-quire no sample preparation. The proprietary MultiPath technology achieves high sensitivity and quan-tification by using digital non-magnified imaging to count fluorescently labeled cellular and molecular targets. The MultiPath Platform eliminates labile enzymatic reagents, biochemical purification, and liq-uid handling (including wash steps), reducing complexity and cost while increasing throughput. We pro-pose to develop high-performance easy-to-use tests for important diagnostic applications for initial commercialization in the hospital setting. The platform and tests will also have commercial and medical advantages in outpatient settings. For Step 2, we plan to: (1) develop MultiPath assays for Clostridium difficile (C. difficile) diagnostics, multi-drug resistant Staphylococcus aureus (MRSA) surveillance, and urinary tract infection (UTI) diagnostics, (2) develop the MultiPath consumable cartridges for the three tests, (3) develop the benchtop automated MultiPath Analyzer, and (4) demonstrate the performance of the integrated MultiPath Analyzer and MultiPath Tests.

Principal Investigator: Ephraim Tsalik
Institution: Duke University, Durham, North Carolina
Executive Summary: Inappropriately prescribed antibacterials for viral respiratory illness contribute to increased healthcare costs, unnecessary drug-related adverse effects, and drive antimicrobial resistance. The inability to rapidly and reliably distinguish bacterial from viral or non-infectious etiologies is a major impediment to appropriate antibiotic use. Pathogen detection strategies can be helpful but are limited by poor sensitivity, long time-to-result, inability to distinguish infection from colonization, or restricted number of target pathogens. Peptide biomarkers such as procalcitonin may also be helpful but are poorly sensitive and specific. Consequently, these approaches have not adequately addressed the antibacterial overuse problem. We therefore propose an innovative solution focusing on the patient’s response to infection. New scientific advances can now capture the entirety of the host response using system-wide molecular surveys (e.g., RNA, proteins, metabolites). We have developed analytical methods to define the stereotyped responses found within these highly complex and dense data. Applying these techniques to infection, we have shown the pattern of immune system response can distinguish bacterial, viral, and non-infectious etiologies. That response is most robustly detected in the patient’s gene expression profile, which is far more accurate than existing diagnostic tests. This strategy is only useful, however, if it can be measured rapidly, simply, and at the point-of-need. Supported by the Antibacterial Resistance Leadership Group, Duke University and BioFire Diagnostics now propose such a test. This simple-to-use, 1-hour test distinguishes bacterial infection, viral infection, or neither so as to guide the appropriate administration of antibacterials at the point-of-need.

Principal Investigator: Joe Frassica
Institution: Philips North America, Cambridge, Massachusetts
Executive Summary: Philips and the University of Pennsylvania (Penn) jointly propose development of a rapid, point of-care system to reliably rule out bacterial infection at primary care within minutes to avoid the unnecessary use of antibiotics. Between 80-90% of all antibiotics are prescribed in a primary care1 setting. Given the current misuse of antibiotics and related antimicrobial resistance, primary care settings can play an important role in reducing unnecessary antibiotic prescriptions2. Despite the benefits of using a point-of-care test to detect infectious disease biomarkers, adoption in primary care remains modest3. The reason for low adoption is that currently available biomarkers have low levels of clinical performance (i.e., sensitivity and specificity) and available tools require advanced skills to operate.

Philips is developing a unique approach to reliably detect bacterial infection at the point of care using a fast and easy-to-use test. The solution is designed to detect the biomarker Human
Neutrophil Lipocalin (HNL) on a diagnostic platform from a single drop of human blood. In 2015 the concept was successfully demonstrated with blood samples from patients with acute fever where in only minutes patients with a bacterial infection could clearly be distinguished from those without. To make this marker available in the primary care setting, Philips is working to realize the whole blood based assay onto its Minicare system. The CE-marked Minicare system is a platform for rapid blood testing, consisting of a point-of-care analyser and cartridges supporting different types of tests. Finalization of product development is estimated at 2.5 - 3 years. Penn will confirm performance of the biomarker using U.S. clinical samples in a separate study. Once test development is complete, Penn will serve as the clinical study site to validate and register the test and to perform a health technology assessment measuring its impact in reducing unnecessary prescription of antibiotics.

Principal Investigator: Gregory Loney
Institution: Click Diagnostics, Inc., San Jose. California
Executive Summary: Click Diagnostics was founded with the mission to dramatically improve health outcomes through innovative new diagnostic technology providing physicians with “patient-side” test results enabling immediate and accurate treatment. Through dramatic miniaturization, optimization, and cost reduction of laboratory polymerase chain reaction (PCR) technology, we have created a fast, inexpensive, easy-to-use, single-use (disposable) diagnostic device. Think “home pregnancy test,” but for nearly “any infectious disease.” Click has made patient-side testing for infectious diseases almost as simple as taking a patient’s blood pressure. We expect this to revolutionize health testing, moving molecular testing from laboratories and large healthcare facilities to doctors’ offices and clinics, and eventually the home. Because the Click device is fundamentally a PCR assay, products can be developed for most common infectious diseases such as sexually transmitted infections (STI), hospital-acquired infections (HAI), and respiratory infections. Our first product for detecting the STIs gonorrhea, chlamydia and trichomoniasis is scheduled for clinical trial in 2017. As part of the AMR challenge, we plan to add a test for drug resistant gonorrhea to this panel. The CDC identifies drug resistant gonorrhea as an urgent threat due to high levels of drug resistance and potential impact on public health.

Principal Investigator: Ravi Kant Verma
Institution: Spectral Platforms, Inc., Monrovia, California
Executive Summary: We propose InSpector-01 (~ a rapid test for characterizing the presence/absence of pathogenic microorganisms in blood and urine samples) and InSpector-02 (~ a rapid test for characterizing the functional MIC minimum inhibitory concentration of a candidate antimicrobial required to suppress the growth of the causative pathogen, even prior to that pathogen being isolated and identified). InSpector-01 would be used to test blood and urine samples, both in the outpatient and in the inpatient setting, and would identify patients for whom antimicrobial therapy is inappropriate. Thus, InSpector-01 can reduce inappropriate antimicrobial use for the vast majority of patients who are currently treated with antimicrobials, but who are uninfected. InSpector-02 would be used to characterize blood samples in the inpatient setting for the functional MIC of the causative pathogen, so as to guide subsequent antimicrobial therapy (from the 2nd antimicrobial dose on). Both InSpector-01 and InSpector-02 offer unique capabilities ~ we are not aware of any other current/emerging methods that can identify patients for whom antimicrobial therapy is inappropriate (as InSpector-01 does) or current/emerging methods that can provide the functional MIC (as InSpector-02 does; other methods provide markers of specific resistance mechanisms). Both InSpector-01 and InSpector-02 are at an advanced stage of development, and are currently being field tested to develop a quality system necessary for distribution.

Principal Investigator: Ken Babock
Institution: Affinity Biosensor, Santa Barbara, California
Executive Summary: The spread of antibiotic resistance can be slowed by reducing the time needed to administer an effective, targeted antibiotic to infected patients. Our response to the AMR Challenge is to produce a test that both confirms bacteremia caused by highly resistant strains, and provides a complete phenotypical antibiotic susceptibility test (AST) in a time that is radically shorter than today’s clinical standard. The test employs a microfluidic sensor that counts and weighs individual microbes at high throughput, and robustly measures bacterial culture growth at a speed limited only by the growth of the microbes themselves. It will provide clinically actionable information - confirming the presence of highly resistant bacteria and identifying the best targeted therapies - in hours or even minutes, vs. the current standard of 2-3 days. The initial targets will be CRE and ESBL-producing enterobacteriaceae, but the test will be extensible to all high-value bacterial targets and any desired antibiotics, and to a variety of bodily sources of infection (e.g., bacteremia/sepsis, spinal and pleural fluid, urinary tract, etc.) Because it assesses growth phenotypically, the test will not be defeated by evolving resistance mechanisms, and will provide an essential complement to resistance biomarkers and organism identification. The outcome will be the fastest possible sample-to result phenotypical susceptibility test, with the potential to migrate from a laboratory/in-patient setting to the doctor’s office.

Principal Investigator: Ann Falsey
Institution: University of Rochester, Rochester, New York
Executive Summary: More accurate tests for microbiologic diagnosis of acute respiratory infections (ARI) are needed. Acute respiratory infections occur commonly throughout life, accounting for substantial morbidity and mortality in adults, and these infections are a leading cause of antibiotic overuse. Unnecessary antibiotic use is a major driver of the increase in antimicrobial resistance, which is considered to be one of the most urgent threats to global public health. In most cases of ARI, the precise microbial etiology is unknown and antibiotics are administered empirically in illness in both inpatient and outpatient settings. Although sensitive molecular diagnostics such as polymerase chain reaction (PCR) allow clinicians to rapidly and accurately diagnose a wide variety of respiratory viruses, their impact on patient management and antibiotic prescription has been modest primarily due to concern about bacterial co-infection. Sensitive and specific diagnostic tests for bacterial lung infection are currently lacking. Clinical parameters such as fever, purulent sputum, white blood cell count and radiographic patterns do not provide sufficient precision to reliably distinguish viral from bacterial infections. Thus, “ruling out” bacterial respiratory infection is extremely difficult, resulting in a default position of prescribing antibiotics to almost all patients hospitalized with respiratory infection and many outpatients as well. This practice results in significant antibiotic overuse, with resultant adverse effects and increased antimicrobial resistance. Recently, serum biomarkers such as procalcitonin have shown some promise as a supplement to clinical judgment in assessing patients with ARI but a need for more accurate tests remains. Gene expression profiling of whole blood represents a powerful new approach for analysis of the host response during infection. Preliminary studies indicate that viruses and bacteria may trigger specific host transcriptional patterns in blood, yielding unique “bio-signatures” that may discriminate viral from bacterial causes of infection. This approach can be used to supplement pathogen detection allowing clinicians to target antibiotic use for those patients that need treatment and thereby reducing unnecessary antibiotic use and resultant antimicrobial resistance. New rapid molecular diagnostics that incorporate detection of common respiratory viruses and bacteria combined with a limited number host genes discriminatory for bacterial and viral infection will be developed. The eventual goal is to simplify testing such that point of care ARI diagnostics can be employed in hospitals, emergency rooms, clinics and physicians’ offices for the maximal impact on patient care and antibiotic prescription.

Principal Investigator: Ellen Foxman
Institution: Yale University, New Haven, Connecticut

Executive Summary: Every year, citizens of the United States experience more than one billion infections of the upper respiratory tract. Both viruses and bacteria can cause these illnesses but only bacteria are susceptible to antibiotics. However, we do not have a simple, rapid, comprehensive diagnostic test to distinguish between viral and bacterial upper respiratory infections. Lack of such tests contributes to the fact that outpatient visits for acute respiratory infection are the most common setting for inappropriate antibiotic prescriptions in the U.S, with about half of these prescriptions (~35 million prescriptions annually) not indicated1. Our recent findings indicate a way to develop a diagnostic test to rapidly distinguish between viral and bacterial causes of respiratory illness while the patient is in the doctor’s office. Unlike current tests, which detect the presence of specific viruses, this approach measures the response of the patient’s airway cells to the gmainfection to indicate whether the body is fighting a viral infection or a bacterial infection. This approach has two major advantages over current approaches: (1) many different viruses and bacteria can cause similar respiratory symptoms, but the only tests currently available are pathogen-specific. Therefore a single test can easily miss the organism causing the patient’s illness, and testing for large panels of pathogens is complex and can be prohibitively expensive; (2) unlike other proposed approaches to distinguish viral and bacterial infections based on the body’s response, our approach does not require a blood sample, but will work with a nose or throat swab. This test has tremendous potential to transform ingrained practices of prescribing antibiotics for non-bacterial illnesses by demonstrating the cause of infection to physician and patient during the initial healthcare encounter, and therefore will provide a powerful tool for promoting antimicrobial stewardship.