Advances in Early Detection and Monitoring of Sepsis: Cellular Diagnostics with Neutrophil CD64 Bruce H. Davis, M.D. Trillium Diagnostics, LLC Brewer, Maine U.S.A.

Severe Sepsis: Comparison With Other Diseases Crit Care Med. 29:1303, † National Center for Health Statistics, § American Cancer Society, *American Heart Association ‡ Angus DC et al. Crit Care Med. 29:1303, AIDS*Colon Breast Cancer § CHF † Severe Sepsis ‡ Cases/100,000 Incidence of Severe SepsisMortality of Severe Sepsis AIDS* Severe Sepsis ‡ AMI † Breast Cancer § Estimated annual 12 million cases with suspected Sepsis in North America, UK, and Europe 18 million cases annually worldwide with 1,400 deaths per day Healthcare cost of sepsis ~$16 billion/year

Infection: Inflammatory response

Laboratory Indicators of Clinical Acute Inflammation Response to Infection/Sepsis Standard of Care Diagnostic Assays of Infection/Sepsis Leukocyte Counts (neutrophilia) - CBC –absolute counts –band counts or left shift (immaturity index) Cultures for suspected infection Sedimentation Rate C-Reactive Protein (CRP) New Generation Assays of Infection/Sepsis Granulocyte or PMN CD64 (Leuko64, Trillium Diagnostics) Cytokine and receptor levels (intracellular or plasma, TNF-a, IL-6 (Septest), IL-12, CD14, CD16, etc.) Procalcitonin plasma levels (Brahms) Endotoxin Activity Assay (Spectral Diagnostics) TREM-1 (Triggering Receptor Expressed on Myeloid cells) LightCycler SeptiFast Test (Roche) Triage Sepsis Panel (CRP, MIP-3, NGal Biosite)

Myeloid Antigen Maturation Sequence From Wood and Borowitz (2006), Henry’s Laboratory Medicine

Scientific Basis for Quantitative PMN CD64 as an Improved Diagnostic Test of Infection/Sepsis PMN CD64 expression is negligible in the healthy state (<2,000 molecules per cell) - low false positive rate PMN CD64 expression becomes elevated under the influence of the inflammatory related cytokines, such as interferon-  (3-4 hrs), G-CSF (4-6 hrs), IL-12 Increased PMN CD64 expression results in enhanced antibody- mediated functional responses (phagocytosis, oxidative burst, bactericidal activity) in PMNs – pathophysiologically significant change during inflammation PMN CD64 becomes elevated in the presence of infection/sepsis –Final cytokine pathway effect at cellular level – barometer of “sickness” –Better performance than cell counts, left shift, and CRP High specificity - PMN CD64 expression is not elevated in: –Malignancy of myeloid cells (CML, MPD, MDS) –Any drug therapy (other than cytokines) –Clinical conditions with localized tissue damage (myocardial ischemia, uncomplicated surgery, and exercise injury) –Pregnancy –Auto-immune disorders (Rheumatoid Arthritis, Systemic Lupus Erythematosis)

PMN CD64: Publications indicating utility as an inflammatory marker in sepsis and infection detection Guyre et al: J Clin Invest 86: , 1990 Davis BH et al: Laboratory Hematology, 1:3-12, 1995 Herrara et al: J. Med. Micro. 12:34, 1996 Quayle JA et al: J Immunol 91: , 1997 Leino et al. Clin Exp Immunol 107:37-43, 1997 Fjaertoft et al, Pediatr Res 45:871-76, 1999 Moallem HJ et al: Scand J Immunol 52:184-89, 2000 Bakke AC et al: Clin Appl Immunol Rev 1:267-75, 2001 Barth E et al: Cytokine 14: , 2001 Qureshi et al, Clin Exp Immunol 125:258, 2001 Fisher et al, Intensive Care Med. 27: , 2001 Hirsch et al, Shock 16: 103-8, 2001 Layseca-Espinosa et al,, Pediatr Allergy Immunol 13: , 2002 Ng et al, Pediatr Res 51: , 2002 Allen E et al: Ann Rheum Dis 61:522-5, 2002 Wagner et al, Shock 19:5-12, 2003 Briggs et al, Lab Hematol 9: , 2003 Ng PC et al, Pediatr Res 56: , 2004 Davis BH, Expert Rev Mol Diag, 5: , 2005 Davis BH and Bigelow NC, Laboratory Hematology, 11: , 2005 Fjaertoft G et al, Acta Paediatr 94: , 2005 Davis BH et al, Arch Path Lab Med, 130(5):654-61, 2006 Matsui et al, J Rheumatol 33(12): , 2006Matsui et al, J Rheumatol 33(12): , 2006 Livaditi et al, Cytokine 36: , 2006Livaditi et al, Cytokine 36: , 2006 Cho et al, Thrombosis Res, in press, 2007Cho et al, Thrombosis Res, in press, 2007 Bhandari et al, Pediatrics, in press, 2007

PMN CD64 Expression in Infection Davis BH et al. Laboratory Hematology, 1:3-12, 1995

Davis BH and Bigelow NC, Laboratory Hematology, 11: , 2005 »160 patient blood samples selected from hospital laboratory based upon blood counter flagging »Assays performed: CBC, manual leukocyte differential counts (H20-A), PMN CD64 by flow cytometry »Retrospective blinded chart review with scoring: –0 = No Infection or inflammation –1 = Localized infection or tissue injury –2 = moderate suspicion for systemic infection and/or tissue injury –3 = Documented sepsis and/or severe tissue injury Methods Comparison of Neutrophil CD64, Manual Myeloid Immaturity Counts, and Automated Hematology Analyzer Flags as Indicators of Infection or Sepsis

PMN CD64 Expression vs. Band Percents & Counts Result: Moderate correlation between PMN CD64 Expression and band% and weaker correlation with immature myeloid fraction Conclusion: Indicates CD64, a measure of cell activation and cytokine functional upregulation, to be a related, but likely independent of left shift Davis BH and Bigelow NC, Laboratory Hematology, 11: , 2005

PMN CD64 Correlates Best with Clinical Sepsis Score Davis BH and Bigelow NC, Laboratory Hematology, 11: , 2005

PMN CD64 Best Predicts Presence of Infection/Sepsis Davis BH and Bigelow NC, Laboratory Hematology, 11: , 2005 Positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), and the negative likelihood ratio (LR-) for clinical evidence of infection/sepsis indicate neutrophil CD64 to have the better diagnostic performance relative to left shift determinants (morphology or flags) and cell counts

Infection/Sepsis Markers in Emergency Room Patients Davis BH et al, Arch Path Lab Med, 130(5):654-61, 2006 »Patients randomly selected from emergency department encounters (N=100) »Assays performed: CBC, band counts, Westergren Sedimentation Rate, C-reactive protein, PMN CD64 »Retrospective blinded chart review with scoring: –0 = No Infection or inflammation –1 = Localized infection or tissue injury –2 = moderate suspicion for systemic infection and/or tissue injury –3 = Documented sepsis and/or severe tissue injury Methods

Infection/Sepsis Markers in Emergency Room Patients Davis BH et al, Arch Pathol Lab Med. 130(5):654-61, 2006

Infection/Sepsis Markers in Emergency Room Patients Davis BH et al, Arch Pathol Lab Med. 130(5):654-61, 2006 Positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), and the negative likelihood ratio (LR-) show neutrophil CD64 to have the best diagnostic performance

CD64 on neutrophils is a sensitive and specific marker for detection of infection in patients with rheumatoid arthritis. CD64 on neutrophils is a sensitive and specific marker for detection of infection in patients with rheumatoid arthritis. Matsui TMatsui T, Ohsumi K, Ozawa N, Shimada K, Sumitomo S, Shimane K, Kawakami M, Nakayama H, Sugii S, Ozawa Y, Tohma S. Ohsumi KOzawa NShimada KSumitomo SShimane KKawakami MNakayama H Sugii SOzawa YTohma S Matsui TOhsumi KOzawa NShimada KSumitomo SShimane KKawakami MNakayama H Sugii SOzawa YTohma S J Rheumatol 33(12): , 2006 Department of Rheumatology, Sagamihara National Hospital, National Hospital Organization, Kanagawa, Japan. OBJECTIVE: In inflammatory diseases, differentiation between infection and disease flares is often clinically difficult because of similar signs and symptoms, such as fever and elevation of inflammatory markers. In rheumatoid arthritis (RA), infection is not only one of the major complications but also one of the frequent causes of death. Use of biologic agents such as tumor necrosis factor-a blockers has been reported to increase the incidence of tuberculosis or opportunistic infections. We examined the utility of CD64 (FcgRI) expressed on neutrophils as a marker for detection of infection complicated with RA. METHODS: We measured the expression level of CD64 per neutrophil quantitatively by flow cytometry in 279 samples from 237 patients with RA with various levels of disease activity or types of infection, and in 52 samples from 36 controls including subjects with infection. RESULTS: CD64 expression was significantly higher among RA patients with infection (median 4156 molecules per neutrophil, interquartile range ) than in those without infection (884, IQR ) (p < or = 0.001). The sensitivity of CD64 on neutrophils for the diagnosis of infection (using a cutoff value of 2000 molecules per cell) was 92.7% and specificity was 96.5%. CD64 expression was not affected by the disease activity of RA or the use of corticosteroids, disease modifying antirheumatic drugs, and biologic agents. CD64 was upregulated in infection by bacteria, viruses, fungi, and mycobacteria. CONCLUSION: Our results suggest that quantitative measurement of CD64 expression on neutrophils can be used as a sensitive and specific marker to detect infection complicating RA.

Neutrophil CD64 expression and serum IL-8: Sensitive early markers of severity and outcome in sepsis Olga Livaditi, Anastasia Kotanidou, Aikaterini Psarra, Ioanna Dimopoulou, Christina Sotiropoulou, Kallirroi Augustatou, Chryssa Papasteriades, Apostolos Armaganidis, Charis Roussos, Stylianos E. Orfanos, and Emmanuel E. Douzinas Department of Critical Care, University of Athens Medical School, Athens, Greece Cytokine, received 20 September 2006; revised 29 December 2006; accepted 7 February Available online 27 March Abstract The aim of the present study was to investigate which biomarker/s reliably assess severity and mortality early in the sepsis process. In 47 critically-ill patients within the 24 h of septic onset, Interleukins (IL)-8, -1β, -6, -10, and -12p70, tumor necrosis factor-α (TNF-α), procalcitonin (PCT) and C-reactive protein (CRP) were measured in serum. Additionally, CD64 expression was measured in neutrophils. In early sepsis, neutrophil CD64 expression and IL-8 levels are the only biomarkers that increased with sepsis severity, differentiating disease stages: sepsis, severe sepsis and septic shock (p < 0.001). The biomarkers that best evaluate the severity of sepsis (via APACHE II) were CD64, IL-8 and IL-6 (p < 0.01), and the severity of organ failure (via SOFA) were CD64 and IL-8 (p < 0.01). CD64 expression and IL-8 levels were associated with mortality within 28-days (OR = 1.3, p = 0.01 for CD64 and OR = 1.26, p = for IL-8 by logistic regression analysis) and ROC curve analysis showed high sensitivity and specificity for predicting sepsis stages and the 28 day mortality. We conclude that there is an early increase of neutrophil CD64 expression and IL-8 levels during sepsis. Based on this single measurement it is possible to reliably assess the stage, detect the severity and predict the 28-day mortality of sepsis.

Leuko64 Assay Kit* from Trillium Diagnostics, LLC Components:  Reagent A: Cocktail of Anti-CD64 FITC and Anti-CD163 PE monoclonal antibodies  Reagent B: 10X Red Cell Lysis Buffer (flow cytometry version only)  Reagent C: Fluorescent microsphere suspension for instrument set-up and quantitation (traceable to NIST SRM 8640)  Software DVD: automated data analysis and reporting; allows for bead value assignment; package insert; MSDS information; instrument acquisition protocols  Two Versions  Flow Cytometry (75 and 250 test sizes)  Hematology Analyzer (Abbott Sapphire and Cell Dyn 4000, 100 test size) *Patent pending, CE marked Distribution by IQ Products in EU

Trillium Diagnostics Leuko64 Assay: Sample Stability – Preanalytical Blood Storage (< 48 hours)

Trillium Diagnostics Leuko64 Assay: Sample Stability – Post-Staining Blood Storage (< 4 hours)

Trillium Diagnostics Leuko64 Assay: Instrument Set-up and Assay Standardization with Beads (reagent C)

Leuko64 Software: Automated cluster finding gating of calibration beads – co-developed with Verity Software House Calibration Bead Gating Purpose of calibration beads: Instrument set-up CD64 Index CD163 Index Lot to lot correlations

Software uses cluster finding algorithms to locate lymphocytes, monocytes, and granulocytes. Lymphocytes serve as an internal negative control (CD64 negative) and software will alert user if lymph gate has a CD64 index greater than 1.00

Monocytes serve as an internal positive control (monocytes normally have moderate levels of CD64) Software will alert user if monocyte gate has a CD64 index less than 3.00, which may indicate need for gate adjustment or failure to add proper antibody volume (reagent A).

PMN CD64 Index reported with values of < 1.00 typical of healthy individuals Software has database function that can imported into Microsoft Excel

PMN CD64 Index reported with values of >1.50 indicative of infection

Leuko64 Software: Summary Report

PMN CD64 as a Sepsis Marker in Neonates Fjaertoft et al Pediatr Res 45:871-76, 1999 Similar reports by: Layseca-Espinosa et al. Pediatr Allergy Immunol, (5): , Ng et al. Pediatr Res, (3): , and Ng et al. Pediatr Res, 2004; 56(5):

Neonatal Study: PMN CD64 Index correlates best with C- Reactive Protein and presence of sepsis Y=1.0404X – r =

Inflammatory Markers in Neonates

Characteristics Healthy Controls (n=55) suspected DIC (n=97) P value IL-6 (pg/ml) 4.7   IL-10 (pg/ml) 7.8    Neutrophil CD     3707 <0.001 <0.001 Elastase (μg/mL) 57.9    Monocyte CD     7.1 <0.001 <0.001 Monocyte CD   hsCRP 0.20    8.0 <0.001 Inflammatory Parameter Comparison between Healthy Controls and Patients with Suspected DIC Data from Dr. Han-Ik Cho, Department of Laboratory Medicine, Seoul National University College of Medicine, Seoul, Korea

CharacteristicsSurvivors(n=69)Non-survivors(n=28) P value IL-6 (pg/ml) 79.3   <0.001 IL-10 (pg/ml) 23.6    Neutrophil CD     4796<0.001 Elastase (μg/mL)    Monocyte CD     Monocyte CD   hsCRP 8.02    Comparison between survivors and non-survivors in patients with suspected DIC Data from Dr. Han-Ik Cho, Department of Laboratory Medicine, Seoul National University College of Medicine, Seoul, Korea

Comparison of neutrophil CD64 and IL-6 according to concomitant infection between non-overt DIC and overt DIC. Data from Dr. Han-Ik Cho, Department of Laboratory Medicine, Seoul National University College of Medicine, Seoul, Korea

Neonatal Sepsis – Yale Study Table 1. Characteristics of the neonatal population. Variables No Sepsis (n=123) Sepsis present (n=40) p value Gestational age (weeks)*   Birth weight (gms)* 1969   Male gender (%) 68 (55) 24/39 (62) Vaginal Delivery (%) 33/120 (28) 13/38 (34) min. Apgar score** 7 (1 - 9) 6 (1 – 9) min. Apgar score** 8 (1 - 9) 9 (3 – 9) Ventilation days* 14.0   NCPAP days* 3.6   Days on Oxygen* 19.1   Length of Stay (days)* 38.7   *Mean  SEM; ** Median (range); NCPAP: nasal continuous positive airways pressure From Bhandari V, Wang C, Rinder C, and Rinder H, submitted for publication, Pediatrics (April, 2007)

Neonatal Sepsis – Yale Study Table 2. Characteristics of the sepsis episodes. Variables No Sepsis (n=165) Sepsis present (n=128) p value Age at sepsis work up (days) * 12.3   Hematocrit (%)* 42.7   White Blood Cell Count/ mm 3 * 14.0   1.2 < Segmented neutrophils (%) * 41.9   Bands (%) * 4.5   0.9 < Platelet count (k/ mm 3 ) *   8.5 < Absolute Neutrophil Count (/ mm 3 )* 6,324  326 8,929  Absolute Band Count (/ mm 3 )* 698  60 2,716  242 < Immature/Total neutrophil ratio * 0.05   0.01 < PMN CD64 index * 2.63   *Mean  SEM “For all sepsis episodes, using ROC analysis, CD64 had an AUC of 0.74; using a cutoff value of 2.30, CD64 in combination with ANC had the highest negative predictive value (93%) for ruling out sepsis and 95% sensitivity for diagnosing sepsis. For confirmed culture-positive sepsis episodes, CD64 had the highest AUC (0.852) of all hematologic variables, with a sensitivity of 80%, and a specificity of 79%, using a cutoff value of 4.02.” From Bhandari V, Wang C, Rinder C, and Rinder H, submitted for publication, Pediatrics (April, 2007)

Neonatal Sepsis – Yale Study ROC curve analysis of hematologic parameters in confirmed sepsis episodes. CD64 had the highest area under the curve (AUC: 0.852). From Bhandari V, Wang C, Rinder C, and Rinder H, submitted for publication, Pediatrics (April, 2007)

CD64 Expression Sample #4 Flow Cytometer Sample #47 Flow Cytometer Sample #3 Flow Cytometer Sample #1 Flow Cytometer Sample #4 CD-4000 Sample #47 CD-4000 Sample #3 CD-4000 Sample #1 CD-4000 CD64 Analysis on Abbott CD-4000: Comparison with Flow Cytometer

Flow Cytometry - FACScan Abbott Cell Dyn Sapphire

Leuko64 Measurements on Abbott Cell Dyn Instruments: Correlation with Flow Cytometry SapphireCD-4000

Anticipated Clinical Utility of Leuko64 Assay of PMN CD64 Expression Screening for infection/sepsis or illness severity in outpatients and hospitalized patients - triage role Therapeutic monitor of antibiotic response in infection –potential indicator for conversion of I.V. to oral therapy –benefit of reduction in antibiotic use and subsequent development of resistant organisms Therapeutic monitor of G-CSF therapy Infection screening of post-operative and post-chemotherapy patients, HIV+ patients, and others at risk for infection/sepsis Distinction between inflammatory leukemoid reaction and myeloproliferative disorder in patients with unexplained neutrophilia Adjunct test with blood cultures –Earlier indicator of patients with sepsis prior to culture result availability –Interpretation of false positive blood cultures with contaminate bacteria

Acknowledgements: Kathleen T Davis - Trillium Nancy C Bigelow – Trillium Karen Becker - Trillium Victoria Kinney and Hematology Lab – MMC Dan Sobel – MMC Neonatology Sam Machin - UCL, U.K. Carol Briggs – UCL, U.K. Harvey Rinder – Yale Univ Paul Guyre - Dartmouth Bob Kisabeth – Mayo Abe Schwartz - CQC Bruce Bagwell – Verity Ben Hunsburger – Verity Don Wright - Abbott Steve Scott – Abbott Richard Kendall - Abbott R&D Systems, Inc. IQ Products, BV