Kantonsspital Aarau Universitätsklinik der Medizinischen Fakultät Basel Die Rolle von Procalcitonin und anderen Biomarkern in der Sepsis ? Beat Müller, M.D. 23 Juni 2011 Why is a critically-ill medical patient “septic”? „Bacterial“ Sepsis ≈ 65% organism known, <50% pos BC „Bacterial“ CAP ≈ 30% organism known, ≈ 10% pos BC ? others Brain CardioVascular Dyspnea (AECB…) Sepsis 33% Pneumonia 67% UTI GIT others Müller B, Crit Care Med 2000; 28: 977-83 Christ-Crain M, AJRCCM, 06; 174: 84-93 „Patientflow“ „Common Cold“ RTI Setting Primary Care Emergency Room Pneumonia Hospital Sepsis ICU Outpatient Diagnosis RTI ? Prevalence Mortality Bronchitis Therapy Antibiotics ? Management Prognosis ? Inpatient ICU EU ≈500Mio <<1% ≈50Mio <1-3% ≈5Mio 5-20% ≈0.05Mio 30-70% >100 infection and sepsis markers proposed Why Procalcitonin? PCT 1 1 Sensitivity Sensitivity CRP IL-6 PCT + Clinical Diagnosis Lactate Clinical Diagnosis alone 0 0 0 1- Specificity 1 0 1- Specificity 1 Müller B, Crit Care Med (2000) 28:977-983 Harbarth S, Am J Respir Crit Care Med (2001) 164:396-402 Risk-adapted Medicine Family Physician ! “Overruling” (“Gatekeeper”) (High-risk, Co-Morbidities) Diagnosis Respiratory Tract Infection? Therapy Antibiotics ? side effects↑ costs↑ antibiotic resistance↑ History & Clinical Exam “Biomarker” (essential!) (Grenzbereiche) Signs & Symptoms Culture, Serology Acute Bronchitis COPD Exacerbation 70% of antibiotics for 70% viral (!) respiratory infections! Radio -logy <0.1 CAP STOP Pneumonia Antibiotics 0.1-0.25 0.26-0.5 Stop/Start based on Algorithm ≥0.5 START Antibiotic PCT guided AB treatment ER LRTI ICU Schütz et al., Exp Rev Anti Infect 2010 Procalcitonin to reduce antibiotic use Schütz et al., Exp Rev Anti Infect 2010 Safety of PCT guided antibiotic stewardship SAE n=233 (17.2%) SAE n=168 (18.1%) Schütz P et al, JAMA 09 Schütz P et al, JAMA 09 PCT-guided AB-Therapy in ICU‘s? Procalcitonin to reduce antibiotic use Schütz et al., Exp Rev Anti Infect 2010 The PRORATA Study Multicenter, ICU, Sepsis N = 621 Antibiotic Duration Outcome 23% more antibiotic free days alive Bouadma et al, Lancet 2010 Summary - Procalcitonin Guided Antibiotic Therapy RTI Setting Mortality „Common Cold“ Primary Care Bronchitis Emergency Room Pneumonia Hospital <<1% <1-3% 5-20% 30-70% 14% 125d 64% 0% 106d AB-Initiation 75% Duration AB exposure 75% 44% 40% 44% 40% The less antibiotic exposure, the less antibiotic resistance! Sepsis ICU 40% Christ-Crain et al., Lancet 04 Christ-Crain et al., AJRCCM 06 & 08 Stolz et al., CHEST 07 Nobre, AJRCCM 07 Briel et al., Arch Int Med 08 Schütz et al., JAMA 09 Stolz et al., ERJ 2010 Bouadma, Lancet 2010 If PCT would be used in „real life“ for RTI... Reduced AB-Prescription using PCT-Guidance 95%? 75% Asia! USA! Latin America! PCT PARTI - study, Briel M., Arch Int Med 2008 Filippini M, Health Policy, 2006 ProREAL-Team: Dr. W. Albrich Dr. M. Batschwaroff Dr. F. Dusemund Dr. P. Schuetz Mrs. K. Regez, RN Mrs. U. Schild, RN Mrs. R. Bossart, RN Principal Investigator: Prof. Dr. B. Müller Study sponsor: bioMérieux, Lyon (F) ProREAL-Contact: +41 79 706 3003 [email protected] Co-Investigators Switzerland Prof. Dr. W. Zimmerli (Liestal) PD Dr. E. Bächli (Uster) Dr. S. Meier, S. Kraljevic (Menziken) Dr. B. Bucher, A. Spillmann (Muri) Prof. Dr. S. Bassetti (Olten) Dr. R. Thomann (Solothurn) Dr. Thomas Sigrist (Klinik Barmelweid) Dr. Daniel Rodrigues (GP, Baden) France Dr. G. Beaune (Annecy) Dr. J. Gaillat (Annecy) Prof. Dr. P. Hausfater (Pitié Salpêtrière, Paris) Dr. P. Banel (Beaume les Dames) Dr. E. Carre (GP‘s association president) Dr. P. Chatron (GP, Clermont-Ferrand) Dr. J. Thierry (GP, Lyon) USA Dr. D. Amin (Clearwater, FL) 1801 patients enrolled . . . ! Good compliance is feasible in real-life conditions and leads to shorter duration of AB therapy 72.4% Compliance 2.3% respiratory instab 0.2% hemodynamic instab 0.2% imminent death 0.8% chronic immunosuppression 0.7% chronic or other infection - 6.0d - 5.7d - 1.7d - 76% - 79% - 18% - 6.1d - 3.8d - 43% 81% 0.5% complication/difficult-totreat-organism 3.9% low PCT, high clinical severity 19.0% no predefined criteria Over entire index presentation: algorithm-compliance 68.2% experience country, treatment site Algorithm-Compliance depends on country, treatment site, experience & diagnosis F, ambulatory 85.1% P=0.63 CH, ambulatory 87.6% F, hospital 66% P<0.0001 USA, hospital 33.5 % P=0.06 P<0.0001 CH, hospital 74.5 % naive 60.1 % P<0.0001 experienced 82.5 % diagnosis Influenza 81.0% CAP 63.7 % COPD exacerbation 70.1 % Bronchitis 0% 81.0% 20% 40% 60% 80% 100% Compliance Limitations of Procalcitonin & Reasons for „Overruling“ Limitations of PCT as a marker of infection • Cut-off range depends on clinical setting – PCT does not replace the doctor („pretest-probability“!) – Co-Morbidities? Setting? Site & Extent of Infection? Assay? • False positives & negative values occur (≈10%) – pos: SIRS, ARDS, „cytokine storm“, malaria, newborns… – neg: early, localised, subacute, immunocompromised, fungal... Procalcitonin remains low in localised infections 75 yrs old patient with Empyema T: 38.1°C, CRP: 150 mg/L, Lc: 13.4x109 PCT: 0.19 ng/ml Procalcitonin for antibiotic stewardship in other sites of infections? Meningitis Sepsis Endocarditis Respiratory tract infections UTI Abdominal infections e.g. Pancreatitis, Diverticulitis Arthritis Osteomyelitis Limitations of PCT as a marker of infection • Cut-off range depends on clinical setting – PCT does not replace the doctor („pretest-probability“!) – Co-Morbidities? Setting? Site & Extent of Infection? Assay? • False positives & negative values occur (≈10%) – pos: SIRS, ARDS, „cytokine storm“, malaria, newborns… – neg: early, localised, subacute, immunocompromised, fungal... • „Single“ PCT measurement is of limited value – Course & prognosis of disease? – Withhold antibiotic therapy? • It cannot identify the pathogen • Cave misuse without therapeutic impact - only then possibly cost-effective Is Procalcitonin testing Cost-effective? Cost: around 24 USD per measurement Heyland KH et al, CCM 2011 What are good Predictors of Bloodculture positivity? Collection of 2x2 blood cultures in 925 patients with CAP 91% negative blood cultures (n=844) 9% positive bood cultures (n=81) 1.00 (84% Strept. pneumoniae) 0.50 0.25 0.00 Sensitivity 0.75 Clinical predictors AUC Age: 0.55 BD systolic: 0.63 Previous AB: 0.59 Temperature: 0.61 0.00 0.25 0.50 0.75 1-Specificity Risk predictors PSI: 0.55 Biomarkers Leukocytes: CRP: PCT 0.57 0.67 0.83 1.00 Müller F. (in preparation) Efficiency versus Economic Benefits of PCT pre-screening PCT cut off (µg/L) Reduction in blood culture collection (%) Total costs (USD) * Missed pathogens Number needed to screen to detect 1 pathogen All patients 0.1 0.25 0.5 1.0 1.5 - 12.6% 36.9% 51.5% 60.5% 67.5% 134’125 117’225 84’633 65’051 52’979 43’591 - 1.2% 3.7% 11.1% 16.0% 18.5% 11.4 9.1 6.5 5.2 4.4 3.5 *Asumming cost for 2 x 2 bottels of blood cultures of 145 USD indirect costs are not included (i.e. nurses work) Müller F. (in preparation) PCT improves antibiotic exposure Does PCT improve length of hospitalisations? Obviously not….but why? Medizin: Strategie vs. Realität proADM KSA Medizin Bettenbelegung 95-105% Pneumonia severity index (PSI) [Fine NEJM 1997;336:243] Co-Morbidität: • Tumor • Herzinsuffizienz • cerebrovask. Erkr. • Nierenerkrankung • Lebererkrankung + + + + + 30 10 10 10 20 Labor & Röntgen art. pH < 7.35 P-Harnstoff > 11mmol/l P-Natrium < 130 mmol/l P-Glucose > 14 mmol/l Hämatokrit < 30% PaO2 <60mmHg (8 KPA) Pleura-Erguss Untersuchung • Verwirrung • HF > 125/min • AF > 30/min • Bdsyst < 90 mmHg • Temp < 35° / >40° + + + + + 20 10 20 20 15 …….……… Punkte < 50 Punkte <70 Punkte 71-90 Punkte 91-130 Punkte > 130 Punkte PSI Klasse --> --> --> --> --> PSI PSI PSI PSI PSI Klasse Klasse Klasse Klasse Klasse I II III IV V Gesamtpunkte…..... Mortalität --> --> --> --> --> ....... ....... + 10 Alter m (in Jahren) f (in Jahren -10) aus Altersheim Mortalität: Mortalität: Mortalität: Mortalität: Mortalität: + + + + + + + 30 20 20 10 10 10 10 …….……… weiteres Management 0.1% 0.6% 0.9% 9.3% 27.0% --> --> --> --> --> Ambulantes Management erwägen Ambulantes Management erwägen Ambulantes Management erwägen Hospitalisation erwägen Hospitalisation / IPS erwägen CAP patients are hospitalised because of (mis)perceived medical reasons Medical factors Inpatient? SAE 5% SAE 25% Nursing Patients` Relatives` Organisational factors preference preference factors Physicians High risk CAP patients (n=215) Outpatient? Percentage (%) Percentage (%) Low risk CAP patients (n=181) Prognostic Risk Assessment! Nurses Patients Medical factors Nursing Patients` Relatives` Organisational factors preference preference factors Relatives Bähni C, Meier S., Spreiter P, BMC Pulm Med 2010 Diagnostic & Prognostic Biomarkers ? MR-proADM Risk-adapted Medicine Family Physician ! “Overruling” “Overruling” (“Gatekeeper”) (High-risk, Co-Morbidities) (Wish of patients & relatives) Diagnosis Respiratory Tract Infection? Therapy Antibiotics ? Management Hospitalisation? History & Clinical Exam Procalcitonin Prognost. Assessment (essential!) (Grenzbereiche) (Scores, Biomarker, Pflege) Signs & Symptoms Culture, Serology Acute Bronchitis COPD Exacerbation Radio -logy <0.1 CAP STOP Pneumonia Antibiotics 0.1-0.25 0.26-0.5 Stop/Start based on Algorithm ≥0.5 START Antibiotic Sehr tief Tief Outpatient Hoch Short stay NLU / Rehab Home-Nursing OPTIMA Sehr hoch Hospital. ICU Outcome prediction in CAP 42 24 p = ns 20 ProCT (ng/ml) Temperatur Temperature (°C) 41 40 39 38 16 12 8 4 37 0 36 I1 II2 3 4 III IV PSI Klasse class PSI 500 Mortalität p =0% ns 1 5 V 35 Leukozten 300 200 100 3 4 5 40 30% 400 2 PSI class Leucocyte count (x109) Protein C-reaktives CRP (mg/L) p <0.001 p = ns 30 25 20 15 10 5 0 I1 II2 3 4 III IV PSIKlasse class PSI V5 0 I1 II2 3 4 5 III IV V PSIKlasse class Christ-Crain et al., Crit Care 06 PSI Stress hormones to predict severity of pneumonia 200 p < 0.001 p < 0.001 Cortisol (nmol/L) Copeptin (pmol/L) 240 200 160 120 80 150 100 50 40 0 I II III IV PSI Klasse V 0 I II III IV PSI Klasse V Mortalität 0% 30% 0% 30% Christ-Crain et al., Eur J Clin Invest 07 Christ-Crain et al., Am J Respir Crit Care Med 07 Cortisol predicts Mortality in CAP 3500 120 P<0.001 P=0.001 110 Free cortisol (nmol/L) Total cortisol (nmol/L) 3000 2500 2000 1500 100 90 80 70 60 1000 50 1.00 Survivors Survivors Non-Survivors Non-Survivors n=85 0.50 0.75 n=166 Total cortisol < 960 nmol/L 0.25 Total cortisol > 960 nmol/L 0.00 Probability of survival P < 0.0001 0 20 40 Analysis time (days) 60 Christ-Crain M, AJRCCM 2007 Adrenomedullin in pneumonia 6 p < 0.001 ProADM (nmol/L) 5 4 3 2 1 p < 0.02 0.50 PSI: 0.75 ProADM: 0.76 Combined: 0.78 0.00 0.25 Sensitivity 0.75 1.00 0 250 0.00 0.25 0.50 1-Specificity 0.75 1.00 Adrenomedullin NH2- QKL QG DNV AP V A K L Q G Y -C F Y T P D R H R N K R C S K IS Q S Q N D T F S M I T GCR G YQ F F O NH2 Adrenomedullin The CALC Gene Family I' II' III' IV' V' VI' I' II' III' IV' V' VI' CT/PCT CGRP I CGRP II I II II' III' VI' Amylin ADM I' V' IV' V' - One of the most potent vasodilating agents - Bactericidal activity - Increased in sepsis, higher in non-survivors than survivors Prognostic Biomarkers in LRTI Prediction of Mortality & SAE (PSI & CURB65: validated only for mortality in CAP) Biomarkers predict both outcomes in CAP & LRTI ProHOSP, LRTI n=1359 ProADM improves the prognostic posttest accuracy of the PSI Score to predict mortality Pre-test Probability CAP PSI I-III ~ 1% CAP overall ~ 7% CAP PSI IV-V ~ 20% ProADM < 1.8 nM LR- 0.28 ProADM > 1.8 nM LR+ 2.9 LR Post-test Probability 40 % ICU ? 20 % 3% 0.2 % Outpatient ? Fagan. NEJM 1975; 293: 257 Hospitalization - more than „just“ medicine The 4 „C“ Community Control „Patient & Relatives“ „Hospital Management“ Cure Care „Doctors“ „Nurses“ Adapted from Glouberman S, Health Care Mgmt Rev 2001 Risk-adapted Patient-Management Observationphase = Recommendation; Interventionphase = Implementation Assessment of medical, nursing & social Risik and need for patient education (empowerment) (with Scores and/or Biomarkers) Biomarker eg proADM RISIK <0.75 0.75-1.5 1.5 >2.5 Low Intermediate High Very high Advice Outpatient Advice • Home-nurse • Nurse led Unit (NLU) • Short Inpatient Advice Inpatient Advice ICU „Overruling“ - Medical factors: C(U)RB65 Score, (PSI, GOLD), schwere medizinische Co-Morbidität - Nursing factors: Barthel Index, Rowland Score - Social factors: Fear, Denial, Opposition, Lack of care at home Follow-up assessment of triage & Rehab-Potential Patients with LRTI 250 / yr Currently (Apr Independent of - Medical assessment 250 - Nursing / functional assessment - Preference of patient & relatives → Avalilability of Beds problematic / erratic Outpatient Home-nursing care ÜbergangsPost-acute pflege Nursing 2010) 250 Nurse-led Unit (NLU) Traditional Hospital (KSA) OPTIMA (Oct 10 – May 11) 250 150 10 Outpatient 30 Home-nursing care 30 Post-acute Care 30 Nurse-led Unit (NLU) KSA «Kosten – Effizienz» oder - Defizienz in der Medizin ? “OPTIMAL” ?? Conclusions Diagnosis of bacterial RTI & Antibiotic Guidance - 11 (!) RCTs from 5 independent groups established Procalcitonin as novel „gold standard“ biomarker for antibiotic guidance in RTI & sepsis. Prognosis & Hospital Management - Substantial medical & socio-economic impact - State-of-the-art clinical scores (PSI, CURB-65, CRB-65) NOT validated for non-pneumonia RTI & non-fatal SAEs relevant for hospitalisation - Biomarkers complement & improve the discriminatory ability of these risk scores for mortality & clinical relevant SAEs (death, ICU-admission, complications) Interdisciplinary Approach needed - Medical (“fear”), nursing, social & community-related factors - Collaboration between institutions & creation of incentives (government, hospital, nursing home, “out-of-hospital” care) Our health care resources are limited ! Thus, they need to be allocated to high-risk patients !! DANKE für Ihre Aufmerksamkeit Limitations of PCT as a marker of infection • Cut-off range depends on clinical setting – PCT does not replace the doctor („pretest-probability“!) – Co-Morbidities? Setting? Site & Extent of Infection? Assay? • False positives & negative values occur (≈10%) – pos: SIRS, ARDS, „cytokine storm“, malaria, newborns… – neg: early, localised, subacute, immunocompromised, fungal... • „Single“ PCT measurement is of limited value – Course & prognosis of disease? – Withhold antibiotic therapy? • It cannot identify the pathogen • Cave misuse without therapeutic impact - only then probably cost-effective • Not for Escalation (broader) Antibiotic therapy High / increasing PCT Bad Sepsis-outcome ESCALATION of Antibiotic Therapy? (do you believe that not enough antibiotics are the problem in septic shock & MOF?) Jensen JU, Crit Care Med 06; 34:2596-602 High / increasing PCT Bad Sepsis-outcome ESCALATION of Antibiotic Therapy? PASS trial Patients: 1200 in ICU Intervention: If PCT >1ng/ml or increasing then - (re)-sampling & imaging - «blind» broadening of AB 1° Endpoint: Reduction in mortality? Power: ARR 7.5% mortality 2° Endpoint: ressource use BUT: escalation of measures defined by the protocol ! Problem: What was tested? Effectiveness of PCT and/or of broadening of AB therapy? Jensen JU, Crit Care Med 06; 34:2596-602 ESCALATION of Antibiosis in worsening ICU-Pat.? The PASS Study Setting: ICU = (understandable) high baseline use of antibiotic therapy Patients: 1200, Danish ICU, high quality of care Admitted for failure of respiration > haemodynamic > «infection» Intervention: If PCT >1ng/ml or increasing then - (re)-sampling & imaging recommended - guideline-based «blind» broadening of «empirical» therapy Research Questions: 1° Endpoint: Does it help to give even more/broader antibiotic therapy? Power: ARR 7.5% mortality (no sepsis study ever reached that goal!) 2° Endpoint: Will you use more ressources? Of course you will (escalation of measures is defined by the protocol) -What was tested? The Effectiveness of PCT and/ or of guideline recommended broadening of antibiotic therapy? -Were the therapeutic measures taken effective? We don’t know ! - Early goal directed therapy randomized by PCT? - PCT Immunoneutralisation ? The ProCAP Study (N = 302) Diagnostic accuracy for Bacteremia 100 0.5 80 0.25ug/L 1.0 ProCT CRP Lc T PSI VAS 60 40 AUC 0.86 AUC 0.69 AUC 0.60 Only 11% pos. BC in ‚bacterial‘ CAP in need for AB! 20 0 0 20 40 60 100-Specificity 80 100 Delay? Sensitivity?? Gold standard ??? Müller B, BMC Inf Dis 2007 PCT as a marker of infection in UTI? • Prospective observational study in 30 women with uncomplicated acute pyelonephritis • Measurement of PCT, CRP at inclusion, at day 4 and every day until normalization (PCT < 0,25, CRP < 10) & at the end of antibiotic treatment • At inclusion: - PCT>0.25ug/L: 37% (median 0.16) - CRP>10 mg/L: 90% (median 96) • At day 5: - PCT<0.25ug/L: 80% - CRP<10 mg/L: 17% • Median days to normalization: PCT: 5 days CRP: 8 days Abstract kindly provided by S.Harbarth, C. Delémont et al. SFMU & SAEM 2011 PCT as a marker of infection in UTI? • → PCT: with this cutoff little help for antibiotic discontinuation in APN due to small portion of patients with abnormal PCT at diagnosis. → CRP: frequently elevated at diagnosis, but long time to normalization → None of these markers useful to guide AB treatment discontinuation in APN Other cutoffs? Cave: PCT>0.25 ≠ CRP>10 Abstract kindly provided by S.Harbarth, C. Delémont et al, SFMU & SAEM 2011 Septic or non-septic Arthritis? PCT as a marker of infection in localized orthopaedic infections? - Retrospective study comparing PCT & CRP in patients with localised non-bacteremic orthopaedic infections - 60 infected patients, tx with AB, surgery to cure infection - Postoperative PCT were rarely elevated despite continued infection; PCT not better than less expensive CRP Uckay et al., Swiss med wkly 2010 Limitations of PCT as a marker of infection • Cut-off range depends on clinical setting – Co-Morbidities? Setting? Site & Extent of Infection? Assay? – PCT does not replace the doctor („pretest-probability“!) • False negative & positive values occur (≈10%) – neg: very early, localised, subacute, immunocompromised, fungal... – pos: SIRS, ARDS, „cytokine storm“, malaria, newborns… PCT to guide duration of AB therapy in surgical ICU patients: a randomized controlled trial Choice and Duration of Antibiotic therapy 110 Surgical ICU patients with suspicion of sepsis Stopp AB if clinical signs improved and PCT <1 ng/ml or the PCT value was >1 ng/ml, but had dropped to 25 to 35% of the initial value over three days. Similar outcomes (SOFA & mortality) Conclusion: Monitoring of PCT is a helpful tool for guiding AB treatment in surgical intensive care patients. Hochreiter et all, CC 2009 CRP for RCTs? • is frequently used in Europe (>> US) - Low cost, easy availability, „historic practice“ • improves the „routine“ diagnosis of LRTI – as compared to signs & symptoms, Temp, WBC – severe bacterial infection is unlikely if CRP<50ug/L (SMW 06) • has, like every biomarker, diagnostic pitfalls – hsCRP = marker of „chronic, subtle inflammation“ helpful for assessment of cardiovascular risk (US) – Less wide „diagnostic range“ (CRP: 102, PCT: 105) – more „unspecific“ increase in SIRS, trauma alone (CCM 00) Not prognostic in severe infections (CAP, sepsis; Crit Care 06) – delayed increase during the course (peak 2-3d, Infection 07) limiting its use during the initial work-up & guidance of AB-duration – hepatic production is largely IL-6 dependent falsely low levels with steroids (COPD!, J Leukoc Biol 02) • may therefore be less safe to guide antibiotic therapy Lots of Refs available upon request! CRP dynamics predict AB appropriateness in VAP - Prospective observational study in 68 patients with suspected VAP; CRP measured on diagnosis & 96 hours after start of AB - Aim: to evaluate value of CRP for antibiotic response CRP ratio (follow-up/baseline) of 0.80 = useful indicator of AB appropriateness - Conclusion: Serial CRP measurements after VAP onset may help to evaluate effectiveness of AB treatment Lisboa et al, Crit Care Med 2008 CRP velocity following Antibiotics - Aim: to study whether CRP on admission and delta CRP following initiation of AB provide a tool for distinguishing CAP from COPD exacerbation - Conclusion: CRP velocity can be used to distinguish CAP from COPD exacerbation → Intervention studies needed Justo et al, Eur J Int Med 2009 CRP for antibiotic stewardship in primary care -42% -51% -66% Cals JWL, et al, BMJ 09; 338: 1374 CRP to reduce AB in the ED? 139 Patients presenting to the ED with acute cough randomized to control or CRP-group control group: recommendations for chest X-ray and AB CRP group: same algorithm plus CRP testing. AB tx Control group: 31%, CRP group 37%, p=ns Gonzales et al., J Emerg Med 2008 Biomarkers for antibiotic stewardship: anything beyond PCT? sTREM-1 suPAR PCT (ug/L) Neutrophils x109 cells/L MIF (ug/L) CRP (mg/L) sTREM-1 (ug/L) suPAR (ug/L) Multimarker-Approach Kofoed et al., Crit Care 2007 Multimarker-Approach AUC‘s for detection of a bacterial cause of inflammation: 0.50 (0.40-0.60) 0.61 (0.52-0.71) 0.63 (0.53-0.72) 0.74 (0.73-0.81) 0.72 (0.63-0.79) 0.81 (0.73-0.86) 0.84 (0.71-0.91) 0.88 (0.81-0.92) Kofoed et al., Crit Care 2007 Combination of PCT & BPW (biphasic transmittance waveform) from the activated partial thromboplastin time (aPTT) for prediction of sepsis in 200 adults admitted to ICU Zakariah et al., Crit Care Med 2008 Combination of CRP and IL-8 to reduce AB AB treatment in IL-8 & CRP group: 36.1% Standard group: 49.6% p<0.001 Similar % of infections missed at initial evaluation (~ 15%) The number of newborns who received postnatal antibiotic therapy could be safely reduced with a diagnostic algorithm that includes measurement of IL-8 & CRP Franz et al., Pediatrics 2004 Conclusion • > 100 different biomarkers for infection & sepsis proposed • RCT‘s very limited ! • Procalcitonin: best & most frequently studied biomarker, 11 RCT‘s! → 35-70% reduction in antibiotic use without apparent negative impact on patient outcome → Tested in primary care & hospital setting (ED, ICU) → limitations: - usefulness in other infections (e.g. UTI)? - too expensive - false positive & negatives occur - lacks necessary accuracy to be used without clinical judgement Anything beyond PCT? • C-reactive protein: proved to be useful for AB stewardship in primary care, probably better marker in localised infections → limitations: • rather unspecific • delayed increase • No RCT‘s in ER, ICU in adults! • New biomarkers? → many tested in observational studies no intervention studies available What should we achieve by 2020? • Test new biomarkers in intervention studies • Combine promising markers & test in intervention studies → challenge to find the most cost-effective markers & to include in daily practice Accurate Diagnosis of Sepsis is Important for Rapid Initiation of Antibiotics On admission Question Intubated & comatose Temperature: 38.6 °C BP 90/55; P 113 Rales in both lungs Really Sepsis? Procalcitonin cut off* <0.5ng/ml >2.0ng/ml Clinical Interpretation Sepsis Very Unlikely Sepsis Very Likely Consequence for physician Other Diagnosis ? Start AB Therapy *Guidelines from the German Sepsis Society Monitoring PCT Gives you Important Information about Patients’ Prognosis Day 3 of ICU stay Still intubated Temperature subfebrile Need for Vasoactiva Question Procalcitonin Prognosis Decrease No Decrease Clinical Interpretation Responding Not Responding Consequence for physician Continue AB Therapy Other Diagnosis? *Guidelines from the German Sepsis Society Dynamics of Biomarkers upon Infection ug/ml h Endotoxin iv Pneumonia - How long to treat? PCT (ng/mL) T1/2 ProCT: 24h 100 (log-linear) CRP (mg/mL) 10 300 250 200 1 150 100 50 0.1 0 1 0.01 2 3 4 5 6 7 0 8 days on ICU "Normal" Reference range J.C.E., 28yrs Antibiotics iv PCT-guided Antibiotic Therapy (mg/L) 100 Antibiotic (AB) use? GPs COPD ER RTI ICU Trauma Septic Shock 10 Severe Sepsis Sepsis Pneumonia Bronchitis COLD 2 1 0.5 0.25 0.1 Assay? YES! YES! Yes YES Yes No No No NO! NO! NO! LUMItest® PCT LIA PCT-Q® ProCT PCT KRYPTOR® / VIDAS ® Diagnosis Healthy 0.01 Christ-Crain M, Eur Resp J, 07 Schuetz P, Curr Opin Crit Care, 07 Time Course PCT & CRP after Heart Surgery adapted from Sponholz C, Crit Care 06; 10:R145 CRP for antibiotic stewardship in primary care - RCT with 258 patients of 32 GP‘s. Randomization to usual care or CRP assistance Primary outcome: antibiotic use after index consultation Secondary outcomes: patient satisfaction, clinical recovery Cals et al., Ann Fam Med 2010; 8: 124-133 PCT guided antibiotic stewardship in “Real Life” • Question: Is the use of PCT safe & efficient outside controlled study conditions? • Quality control Survey in patients with LRTI in different hospitals • Switzerland (11 hospitals), France (3), USA (1) • Patients with LRTI included in registry via website, where the recommended algorithm was displayed • 1°endpoint: AB duration within 30 days 2° endpoint: adherence to algorithm • 1810 patients enrolled Albrich et al, in preparation PCT guided antibiotic stewardship in “Real Life” • Overall compliance with algorithm on presentation 72.4%, + 8.6% overrulings based on prespecified criteria = 81.0% • Most important overruling criteria high clinical severity • Antibiotic duration significantly shorter in LRTI if PCT algorithm was followed (6.2 vs 8.4 days) Albrich et al, in preparation PCT guided antibiotic stewardship in “Real Life” • Duration of antibiotic treatment was significantly lower as compared to“pre-PCT” time & similar to PCT groups in studies • 18% of overrulings based on prespecified criteria • 10% of overrulings without prespecified criteria Overruled with prespecified reason Immuno-suppression ICU admission High risk patients 2% 2% Anticipated complications Overruled without prespecified reason Decision of the treating physician Not overruled Schuetz P et al, Eur J of Clin Microbiology and Inf Disease, 2010 Outcome and complications Outcome (in LRTI) (no.(%)) In-hospital mortality 79 (5.2%) Need for ICU admission 159 (10.5%) In-hospital overall complications 293 (19.3%) 30-day-mortality 109/1425 (7.6%) 30-day-recurrence rate 32/1425 (2.2%) Risk factors for complications 30 day complications OR 95% CI P CURB65 1.35 1.07-1.72 0.01 CAP (vs. bronchitis) 6.81 2.97-15.61 <0.0001 Multilobar pneumonia 5.25 1.62-17.02 0.006 History of stroke 0.25 0.08-0.77 0.015 Compliance with the algorithm was not associated with risk of complication (p=0.26) The PRORATA Study – Subgroups Antibiotic Duration Outcome Bouadma et al, Lancet 2010 PCT-guided AB-Therapy in Sepsis on the ICU Shorter Antibiotic Duration 6d 10d Nobre V et al, AJRCCM 07 2 Confessions As a clinician, I used to diagnose bacterial infections and prescribe antibiotics easily & accurately. Now, after 20 yrs of research, I am no longer so ignorant. I received funding from vendors of biomarker assays. I never received funding from companies selling antibiotics. So, yes, I might be biased.