Indications for Antibiotics in Exacerbations of COPD Sanjay Sethi MD Professor Pulmonary, Critical Care and Sleep Medicine University at Buffalo, SUNY

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Soler-Cataluña JJ et al. Thorax. 2005;64: COPD Exacerbations: Survival Time (months) Probability of surviving p< p<0.001 p=0.07 3–4 exacerbations 1–2 exacerbations No exacerbation

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Outcome of Exacerbations In ICU patients In ICU patients In-Hospital mortality % In-Hospital mortality % In hospitalized patients In hospitalized patients Hospital mortality 6 - 8% Hospital mortality 6 - 8% In ER patients In ER patients Relapse (repeat ER visit) % Relapse (repeat ER visit) % In outpatients In outpatients Treatment failure rate % Treatment failure rate % Hospitalization rate in treatment failures 16-52% Hospitalization rate in treatment failures 16-52% Connors AJRCCM 1996, Seneff JAMA 1995, Esteban JAMA 2002, Groenewegen Chest 2003, Martin Chest 1992, Murata Ann Emerg Med 1991, Aaron NEJM 2003, Adams Chest 2000, Miravittles ERJ 2001, Ball QJM 1995, Dewan Chest 2000

Antibiotics in AECOPD: Clinical Resolution Anthonisen et al, Ann Intern Med. 1987:106: Spontaneous Resolution at 3 weeks

Antibiotics in AECOPD Clinical Deterioration Anthonisen et al, Ann Intern Med. 1987:106:

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Evidence for Bacterial Etiology of AECOPD Bacteria can be cultured from the distal airways in significant concentrations in >50% of patients Bacteria can be cultured from the distal airways in significant concentrations in >50% of patients Acquisition of strains of bacteria new to the patient is associated with a greater than 2 fold increase in the risk of exacerbation Acquisition of strains of bacteria new to the patient is associated with a greater than 2 fold increase in the risk of exacerbation Monso E, et al. AJRCCM. 1995;152: ; Sethi S, et al. NEJM. 2002; 347; Sethi S, et al. AJRCCM. 2004;168:448-53; Sethi S, et al. Chest. 2000;118:

Evidence for Bacterial Etiology of AECOPD Specific immune responses develop to infecting bacterial strains following exacerbation Specific immune responses develop to infecting bacterial strains following exacerbation Neutrophilic airway inflammation is associated with recovery of bacterial pathogens during an exacerbation Neutrophilic airway inflammation is associated with recovery of bacterial pathogens during an exacerbation Monso E, et al. AJRCCM. 1995;152: ; Sethi S, et al. NEJM. 2002; 347; Sethi S, et al. AJRCCM. 2004;168:448-53; Sethi S, et al. Chest. 2000;118:

Proof of Global Warming

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Efficacy of Antibiotics and Steroids in AECOPD: Systematic Analyses Antibiotics (n=11) Steroids (n=10) OutcomeRRn NNT or NNH RRn Mortality 0.23 ( ) ( ) 9 Treatment Failure 0.75 ( ) ( ) 99 Adverse Effects 2.91 ( ) ( ) 76 Antibiotics + Sputum purulence resolution -- PEFR and gas exchange Steroids + PEFR, FEV 1 and gas exchange Ram FSF et al, Cochrane Lib Vol 2, 2006 Wood-Baker RR et al Cochrane Lib Vol 2, 2006

Anthonisen et al, Ann Intern Med. 1987:106: Sachs et al, Thorax 1995;50: p = ns AECOPD trials: effect of patient selection

AECB trials: effect of patient selection Anthonisen et al, Ann Intern Med 1987;106: Sachs et al, Thorax 1995;50:758-63

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Exacerbation severity is easy to define Exacerbation severity is easy to define Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Antibiotic comparison trials in AECOPD Obaji and Sethi, Drugs and Aging 2001; 18:1-11

Antibiotic trials in AECOPD: Pitfalls Limitation Small n Small n Mild underlying COPD Mild underlying COPD Non-bacterial exacerbations included Non-bacterial exacerbations included End-points compared at 3 weeks after onset End-points compared at 3 weeks after onset Potential consequence › Type 2 statistical error › Diminished perceived antibiotic efficacy › Type 2 statistical error › Spontaneous resolution mitigates differences › Clinically irrelevant Sethi S. Proc Am Thorac Soc. 2004;1:109-14

Antibiotic trials in AECOPD: Pitfalls Limitation Speed of resolution not measured Speed of resolution not measured Lack of long-term follow up Lack of long-term follow up Antibiotic with limited in vitro efficacy Antibiotic with limited in vitro efficacy Poor penetration in to respiratory tissues Poor penetration in to respiratory tissues Potential consequence › Clinically relevant end- point not assessed › Time to next exacerbation not assessed › Diminished perceived efficacy of antibiotics Sethi S. Proc Am Thorac Soc. 2004;1:109-14

Proposed Goals for Treatment of Exacerbations Clinical Faster resolution of symptoms Faster resolution of symptoms Clinical Resolution to Baseline Clinical Resolution to Baseline Prevention of Relapse Prevention of Relapse Increasing exacerbation-free interval Increasing exacerbation-free interval Preservation of health related quality of life Preservation of health related quality of lifeBiological Bacterial eradication Bacterial eradication Resolution of airway inflammation Resolution of airway inflammation Resolution of systemic inflammation Resolution of systemic inflammation Restoration of lung function to baseline Restoration of lung function to baseline Preservation of lung function Preservation of lung function

Bacterial Persistence and Airway Inflammation following AECOPD White et al. Thorax 2003;58: LTB4 (nM) Bacteria eradicated by day 10 Bacteria persisting at day 10 p<0.001 Day MPO (units/ml) Bacteria eradicated by day 10 Bacteria persisting at day 10 p<0.05 p<0.001 Day

MOSAIC Study: Time to First Occurence of Composite Event* ITT population, N=730 *Failure, next AECB or need for further antimicrobial treatment Patients not experiencing composite event (%) Time since randomisation (months) p= Moxifloxacin Comparator Wilson R et al., Chest 2004, 125:

GemifloxacinClarithromycin % patients P = GLOBE : Percentage of Patients with no Recurrences at 26 Weeks Wilson et al., Clin Ther 2002, 24:639-52

Rate of Recovery Antibiotic Choice RR for Slow Recovery RR for Slow Recovery Moxifloxacin vs Clarithromycin 0.41 ( ) Moxifloxacin vs Amox-clav 0.34 ( ) p< Miravittles et al, Resp Med 2005; 99:955-65

Antibiotic Therapy of AECOPD Stratification approach Stratification approach Choose antibiotics based on Choose antibiotics based on Severity of acute illness Severity of acute illness Expected outcome Expected outcome Expected resistance Expected resistance

Proposed Therapies for AECB According to Patient Subsets <4 exacerbations/year No comorbid illness FEV 1 >50% >4 exacerbations/year Serious comorbid illness FEV 1 <50% Home oxygen Chronic oral steroids Recent antibiotic therapy Advanced macrolide Selected cephalosporins Doxycycline TMP/SMX New fluoroquinolones Amoxicillin–clavulanate Fluoroquinolone with antipseudomonal activity (e.g. ciprofloxacin) Simple, uncomplicated AECB Complicated AECB Complicated AECB at risk for P. aeruginosa O’Donnell DE, et al. Can Respir J 2003 Patients with chronic bronchial sepsis Need for chronic corticosteroid therapy and frequent (>4/year) courses of antibiotics FEV 1 <35%

Risk Stratification and Acute Exacerbations of COPD Exacerbations No antibiotics Simple COPD Complicated COPD Cephalosporin (cefuroxime, cefpodoxime, cefdinir), Ketolide (telithromycin), Advanced macrolide (azithromycin, clarithromycin), Doxycycline, TMP/SMX Worsening clinical status or inadequate response in 72 hrs Reevaluate Consider sputum culture MODERATE OR SEVERE At least 2 of the 3 cardinal symptoms: Increased dyspnea Increased sputum volume Increased sputum purulence MILD Only 1 of the 3 cardinal symptoms: Increased dyspnea Increased sputum volume Increased sputum purulence Fluoroquinolone (moxifloxacin, gemifloxacin, levofloxacin), Amoxicillin-clavulanate If at risk for Pseudomonas, consider ciprofloxacin and obtain sputum culture Sethi S, Murphy TF. Infect Dis Clin N Am. 2004;18: Always ask about antibiotic use in previous 3 months

Pathogenesis of Exacerbations Chronic bacterial colonization Chronic inflammation (bacterial + host mediated inflammatory factors) Damaged respiratory epithelium Impaired host defenses: H respiratory virus H new strains of bacteria H environmental irritants Acute on chronic inflammation (bacterial + host mediated inflammatory factors) Progressive loss of lung function and deteriorating quality of life Smoking/Irritants Chronic cycle Acute cycle Antibiotics

Antibiotics: Antibacterial mechanisms Chronic bacterial colonization Chronic inflammation (bacterial + host mediated inflammatory factors) Damaged respiratory epithelium Impaired host defenses: H respiratory virus H new strains of bacteria H environmental irritants Acute on chronic inflammation (bacterial + host mediated inflammatory factors) Suppressive Abx therapy X X Prevent AECOPD X

Trial Overview Mod-severe CB stable phase Moxi 400mg OD x 5 days Screened & Randomized Primary endpoint: no. of exacerbations Placebo OD x 5 days Pulse #2 Pulse #2 8 wks Pulse #6 Pulse #6 8 wks ET 8 wks ET FU #1 8 wks FU #1 FU #3 FU #3 Secondary endpoints: no. of exacerbations diff in lung function HEOR QoL, etc. 48 week treatment period24 week follow-up period N=1132

Conclusions Are Antibiotics Important in the Treatment of AECOPD? Are Antibiotics Important in the Treatment of AECOPD? Moderate to Severe exacerbations: Yes Moderate to Severe exacerbations: Yes Mild: Maybe Mild: Maybe Further Research Further Research Additional Benefit with systemic corticosteroids in moderate exacerbations Additional Benefit with systemic corticosteroids in moderate exacerbations Benefit in mild exacerbations Benefit in mild exacerbations Non-traditional clinical outcomes Non-traditional clinical outcomes Biological consequences of bacteriologic eradication Biological consequences of bacteriologic eradication

Recent Guideline Recommendations for Antibiotic Therapy in AE COPD ATS/ERS —“ May be initiated in patients with altered sputum characteristics ” ATS/ERS —“ May be initiated in patients with altered sputum characteristics ” CTS —“… antibiotics should only be considered for use in patients with purulent exacerbations ” CTS —“… antibiotics should only be considered for use in patients with purulent exacerbations ” ERS — Anthonisen I, Anthonisen II with sputum purulence, severe AE COPD ERS — Anthonisen I, Anthonisen II with sputum purulence, severe AE COPD GOLD —“ Antibiotics are only effective … with worsening dyspnea and cough … also have increased sputum volume and purulence ” GOLD —“ Antibiotics are only effective … with worsening dyspnea and cough … also have increased sputum volume and purulence ” NICE —“ Antibiotics should be used to treat exacerbations of COPD associated with a history of more purulent sputum ” NICE —“ Antibiotics should be used to treat exacerbations of COPD associated with a history of more purulent sputum ” Martinez et al. Expert Rev Anti Infect Ther. 2006;4: (A).

Exacerbation and Health Status SGRQ score Further exacerbation No further exacerbation Spencer et al, Thorax 2003

012 Sustained Quit Intermittent Smoker Continuing Smoker Mean annual change in FEV 1 (ml/year) Mean number of physician visits/year for LRIs Kanner. AJRCCM 2001; 164:358 Decline of FEV 1 by Smoking Status Lung Health Study

Myths in AECOPD Exacerbations are harmless Exacerbations are harmless Exacerbations resolve spontaneously Exacerbations resolve spontaneously Exacerbations are not bacterial in origin Exacerbations are not bacterial in origin Benefits of antibiotics in AECOPD are unproven Benefits of antibiotics in AECOPD are unproven Choice of antibiotics does not matter in AECOPD Choice of antibiotics does not matter in AECOPD

Exacerbation severity Severity of exacerbation = Underlying disease severity + severity of acute episode Severity of exacerbation = Underlying disease severity + severity of acute episode Different classifications Different classifications Anthonisen classification Anthonisen classification Site of treatment Site of treatment Intensity of treatment Intensity of treatment Burge/Wedzicha classification Burge/Wedzicha classification None are validated None are validated

Corticosteroids in AECOPD 147 ED patients 147 ED patients 10 days of 40mg qd oral prednisone or placebo 10 days of 40mg qd oral prednisone or placebo Oral antibiotics: TMP/SMX or Doxy Oral antibiotics: TMP/SMX or Doxy FEV 1 mean 38.3% FEV 1 mean 38.3% Primary endpoint: Relapse in 30 days Primary endpoint: Relapse in 30 days Aaron et al NEJM 2003;348:

Antibiotics in AECOPD: Clinical Resolution Anthonisen et al, Ann Intern Med. 1987:106: Spontaneous Resolution at 3 weeks

Resolution of AECOPD Sethi S, MTSU 2005

Conventional Clinical End-points Large n powering for a 10% difference Large n powering for a 10% difference >1000 >1000 Show a 10% difference Show a 10% difference So what? So what? What does the patient really care about? What does the patient really care about? Getting back to baseline Getting back to baseline Getting there faster Getting there faster Staying at baseline for longer Staying at baseline for longer Avoid complications Avoid complications