Observational Study of Sepsis and Pneumonia to Develop Diagnostic Tests

3 interdependent aims are proposed to discover and initiate development of novel, in vitro
diagnostic tests (IVD) for severe sepsis (SS) and community acquired pneumonia (CAP).

Specific Aim 1: Discovery and initial development of an IVD for early diagnosis of severe
sepsis.

In patients with suspected sepsis, early, accurate identification of patients who will
develop organ dysfunction (SS) is critical for effective management and positive outcome.
While the American College of Chest Physicians/Society of Critical Care Medicine Consensus
Conference definitions provide a clinical guide to identifying patients who have SS, we
propose to develop a rapid, point-of-care (POC) IVD for early diagnosis of SS. The basis of
the proposed IVD will be the measurement of several, host response, plasma proteins. When
performed in patients at the earliest stage of sepsis, this test will have prognostic value,
rapidly identifying patients who will have poor outcomes or complicated courses.

Availability of this IVD will enable patient selection for appropriate disposition, such as
the intensive care unit (ICU), and use of medical therapies, such as early goal-directed
therapy (EGDT), thereby reducing mortality and increasing effectiveness of resource
allocation. A considerable literature exists of host plasma protein changes during sepsis.
Furthermore, in preliminary studies measuring more than 100 host proteins in blood of over
300 patients with SS, we have identified a number of candidate biomarkers of SS. We propose
to inventory, replicate and validate the utility of these biomarkers of SS, and to identify
novel plasma biomarkers of SS, through literature review and a prospective clinical study
employing 2 proteomic technologies (mass spectrometry and multiplexed immunoassays), mass
spectrometry-based plasma metabolomics and sequencing of mRNA derived from peripheral blood
lymphocytes. We intend to enroll 1200 patients with sepsis (evidence of infection and 2 or
more criteria of the systemic inflammatory response syndrome, SIRS) at 3 US tertiary care
hospitals and emergency departments (ED), and to monitor their course both by established
clinical severity indices (Acute Physiology and Chronic Health Evaluation [APACHE II] and
Pneumonia Patient Outcomes Research Team [PORT]scores, and metabolic endpoints such as
lactate, base deficit and pH) and ascertainment of complicating events (such as SS, septic
shock, acute renal failure (ARF), acute respiratory distress syndrome (ARDS),disseminated
intravascular coagulopathy (DIC) and death). It is anticipated that approximately 60% of the
patients will develop SS.

Data will be stored in an anonymized, encrypted, web-based patient registry. Bivariable
analyses will be performed to identify and validate biomarker differences between groups.
Furthermore, we intend to perform predictive modeling using multivariable analyses of the
validated biomarkers and derive a biomarker panel and algorithm for early diagnosis of SS.
The predictive value of the biomarker panel for early diagnosis of SS will be compared with
established prognostic indices, such as metabolic endpoints and APACHE II score. Novel
biomarkers of severe CAP will be identified by mass spectrometry of patient EDTA plasma
samples. Subject to availability of multiplexed immunoassays, some of these biomarkers will
be replicated by immunoassay in the same samples.

During the period of award, a plan for IVD development of the biomarker panel for early
diagnosis of SS will be developed. This is anticipated to involve assay optimization and
transfer to an existing, validated IVD platform, FDA-regulated IVD development processes,
and incorporation of the IVD into an intensive treatment algorithm. The proposed IVD will be
an oligoplex assay performed on a single blood sample using immunoassays on an established
diagnostic platform with time-to-first result of less than 30 minutes and capable of use in
a POC setting, such as an ED or ICU.

Specific Aim 2: Biomarker development for early differentiation of poor outcome in CAP
Complications of CAP, including respiratory failure, other organ system failure and severe
sepsis, are major determinants of morbidity and mortality. At time of presentation with CAP,
accurate identification of patients who will have a complicated course or poor outcome is
critical for effective management and positive outcome. In parallel with Specific Aim 1, we
propose to identify biomarkers for early diagnosis/prognosis of poor outcome in patients
with CAP ("severe CAP"). The biomarkers will be several, host response, plasma proteins that
differentiate mild and severe CAP. Early diagnosis of severe CAP will enable patient
selection for hospitalization, thereby reducing mortality and increasing effectiveness of
resource allocation.

It is anticipated that approximately 33% of the patients enrolled in the Specific Aim 1
clinical study (evidence of infection and two or more SIRS criteria) will have CAP as the
underlying infection causing sepsis. Furthermore, it is anticipated that approximately 25%
of these CAP patients will develop severe CAP. Specific aim 2 proposes a secondary, separate
analysis of all patients enrolled in the Specific Aim 1 clinical study who have CAP in order
to identify biomarkers for early diagnosis of severe CAP. We propose to inventory existing
candidate biomarkers of severe CAP through literature review. Furthermore, we propose to
validate the utility of some of these biomarkers, and to identify a number of novel
biomarkers of severe CAP through analysis of the subset of patients in the prospective
clinical study who have CAP and employing 2 proteomic technologies (mass spectrometry and
multiplexed immunoassays), mass spectrometry-based plasma metabolomics and sequencing of
mRNA derived from peripheral blood lymphocytes. Bivariable analyses will be performed to
identify plasma biomarker differences between mild and severe CAP. Multivariable analyses
will be performed in order to derive a plasma biomarker panel and algorithm for early
diagnosis of severe CAP. The biomarker panel for early diagnosis of severe CAP will be
compared with established prognostic indices, such as PORT score. Novel biomarkers of severe
CAP will be identified by mass spectrometry of patient EDTA plasma samples.

Subject to availability of multiplexed immunoassays, some of these biomarkers will be
replicated by immunoassay in the same samples.

During the period of award, a plan for panel validation and IVD development for early
diagnosis of severe CAP will be developed. The latter is anticipated to involve assay
optimization and transfer to an existing, validated IVD platform, FDA regulated IVD
development processes, and incorporation of the IVD into an intensive treatment algorithm.
The proposed IVD will be an oligoplex assay performed on a single blood sample using
immunoassays on an established diagnostic platform with time-to-first result of less than 30
minutes and capable of use in a POC setting, such as an ED or ICU. This is anticipated to be
a product line extension of the SS IVD.

Specific Aim 3: Biomarker development for early differentiation of sepsis and CAP pathogens
Currently, initial antimicrobial treatment of sepsis and CAP is empiric. Common etiologic
agents in sepsis are gram-positive bacteria (Staphylococcus spp. and Streptococcus spp.),
gram-negative bacteria (e.g., Escherichia coli, Klebsiella spp., and Enterobacter spp.), and
fungi (Candida spp.). Common etiologic agents in CAP are Streptococcus pneumoniae,
Legionella pneumophila, Mycoplasma spp., and viruses. The ability to distinguish these
pathogens at time of presentation of sepsis or CAP would potentially allow more targeted
rather than broad-spectrum initial therapy. Earlier administration of appropriate
antimicrobials would lower patient management cost associated with ineffective therapy and
lessen likelihood of antibiotic resistance. We propose to identify host biomarkers for early
differentiation of up to 4 common etiologic agents in sepsis and CAP. Our preliminary
studies have established proof-of-concept for differentiation between classes of pathogens
in sepsis based on specific differences in soluble host proteins in a blood sample. Based on
our preliminary studies, it is anticipated that approximately 25% of patients in the
Specific Aim 1 clinical study will have a positive blood culture. At least 33 of these
patients are anticipated to have S. aureus bacteremia and 20 gram negative bacteremia.
Specific Aim 3 proposes to compare plasma samples from patients with S. aureus and gram
negative bacteremia in order to identify host biomarkers for early differentiation of
specific class agent in sepsis. As in specific aims 1 and 2, bivariable and multivariable
analyses of biomarkers is proposed to develop a biomarker panel for early differentiation of
staphylococcal and gram-negative sepsis. Similar analysis is proposed to differentiate CAP
pathogens. However, given the absence of a high-sensitivity, gold-standard method for
determination of causal pathogen in CAP, Specific Aim 3 proposes the more conservative goal
of differentiating pneumococcal CAP from atypical CAP based on quantitative differences in
host blood biomarkers. The pneumococcal CAP group will be selected from the clinical
studydataset based on rigorous criteria: S. pneumoniae from blood or sputum culture or
detection of pneumococcal antigen in urine, clinical evidence of CAP and typical (lobar
consolidation) chest radiograph. The nonpneumococcal CAP group will be determined by
negative pneumococcal cultures and urine antigen, clinical evidence of CAP, and an atypical
chest radiograph. It is anticipated that at least 20 patients (15% of the 133 with CAP) will
have confirmed pneumococcal CAP and 40 patients (30%) atypical, non-pneumococcal CAP.
Biomarkers for differentiation of i. S. aureus bacteremia from gram-negative bacteremia, and
ii. pneumococcal CAP from atypical CAP, will be identified by mass spectrometry of patient
EDTA plasma samples. Subject to availability of multiplexed immunoassays, some of these
biomarkers will be replicated by immunoassay in the same samples. It should be noted that
given budget-imposed reduction in patient enrollment of one third from that originally
proposed, we are uncertain that sufficient patients will be enrolled for all Specific Aim 3
analyses to be meaningful. We propose to evaluate the group sizes of enrolled patients by
specific class agent in order to select two specific comparisons between sepsis and CAP
pathogens that are of sufficient size to permit meaningful analysis.

Validation and development of these biomarkers into biomarker panels and rapid, POC, IVD for
early differentiation of pathogen in sepsis and CAP is intended, but is beyond the scope of
the present proposal. A product line extension of the SS IVD is envisaged. Like the test for
early diagnosis of SS, the IVD(s) for early differentiation of CAP and sepsis pathogens will
be oligoplex assay(s) performed on single blood sample(s) using immunoassays or other
analyte assays.

Inclusion Criteria:

1. Patient has known or acute infection or suspected infection AND patient must meet at
least 2 of the following 4 criteria to be enrolled

1. A core temperature of >= 38°C (100.4°F) or <= 36°C (96.8°F)

2. Patients > 18 years of age, Heart rate of >= 90 beats/min Patients 13-18 years of
age, Heart rate of >= 110 beats/min Patients 6-12 years of age, Heart rate of >= 130
beats/min

3. Patients > 18 years of age, Respiratory rate of >= 20 breaths/min Patients 13-18
years of age, Respiratory rate of >= 14 breaths/min Patients 6-12 years of age,
Respiratory rate of >= 18 breaths/min OR PaCO2 of <= 32 mm Hg OR Use of Mechanical
Ventilation for an acute respiratory process

4. Patients > 18 years of age, White cell count >= 12,000/mm3 or <= 4,000/mm3 Patients
13-18 years of age, White cell count >= 11,000/mm3 or <= 4,500/mm3 Patients 6-12
years of age, White cell count >= 13,500/mm3 or <= 4,500/mm3 OR A differential count
showing > 10% immature neutrophils

Exclusion Criteria:

1. Patient is less than 6 years of age.

2. Patient is not expected to survive 28 days because of uncorrectable medical condition
(apart from pneumonia or sepsis), such as poorly controlled neoplasm or other
end-stage disease, or patient has active DNR order

3. Human immunodeficiency virus (HIV) infection with a last known CD4 count of <50 mm3

4. Acute presence of a cerebral vascular event, active gastrointestinal hemorrhage,
seizure (acute episode), drug overdose, burn injury, trauma

5. Patient is pregnant