Does susceptibility testing have a role in predicting

1
Does susceptibility testing have a role in
predicting clinical or microbiological
outcome? Alasdair MacGowan North Bristol NHS
Trust University of Bristol Southmead
Hospital Bristol, UK
2
The paradigm pharmacodynamic microbiological cl
inical index size outcome
outcome Cmax/MIC AUC/MIC TgtMIC
3
The Pharmacodynamic world - students view
AUC/MIC fluoroquinolones daptomycin aminoglycosid
es ketolides metronidazole tetracyclines glycopept
ides oxazolidinones
Cmax/MIC fluoroquinolones aminoglycosides daptomyc
in (metronidazole)
T gt MIC macrolides clindamycin oxazolidinones B
lactams glycopeptides
4
The Pharmacodynamic world - shades of grey
AUC/MIC tetracyclines ketolides
AUC/MIC Cmax/MIC fluoroquinolones aminoglycosides
(daptomycin)
Cmax/MIC
Oxazolidinones glycopeptides
? Dalbavancin
T gt MIC Blactams erythromycin clindamycin
5
Clinical studies showing a relationship between
pharmacodynamic index and outcome
6
Pharmacokinetics and susceptibility
7
Which is dominant - pharmacokinetics or
susceptibility?
8
Pharmacokinetic variability
9
Susceptibility variability - MICs (www.bsacsurv.or
g www.EUCAST.org) Wild type (EUCAST)
Variation ciprofloxacin - E. coli 0.002-0.06
mg/L x30 levofloxacin - E. coli 0.002-0.06
mg/L x30 moxifloxacin - E. coli 0.008-0.25
mg/L x30 Present (BSAC) ciprofloxacin - E.
coli 0.002 - gt 512 gt250,000
10
Which is dominant - pharmacokinetics or
susceptibility? Usually susceptibility drives
changes in pD index hence in situations where
MIC ranges are large, categorical sensitivity
testing should be predictive.
11
Antibiotic resistance and bacteraemia (1) General
cases Phillips et al, 1990 St Thomas Hospital,
London bacteraemias 1969-88, retrospective
analysis Staphylococci, Enterococci,
Enterobacteriaceae, P. aeruginosa Outcomes
12
Antibiotic resistance and bacteraemia
(2) Behrendit et al, 1999 Department of
Medicine, University Hospital, Frankfurt 1989-93,
retrospective analysis Outcome 28d mortality
13
Antibiotic resistance and bacteraemia
(3) Specific settings - ICU Ibrahim et al,
2000 St Louis USA, Medical Surgical ICU (37
beds) 1997-99 Prospective cohort study
Multiple logistic regression- inadequate
antimicrobial therapy as independent determinant
of mortality RR 6.9 (5.1 - 9.3, p lt
0.001) Commonest resistant isolates - VRE,
Candida sp, MRSA, CONS, P. aeruginosa - also
highest mortality
14
Antibiotic resistance and bacteraemia
(4) Specific settings - ICU Harbarth et al,
2002 Geneva, Switzerland, Surgical ICU (22 beds)
1994-7 retrospective cohort study of 244
bacteraemias In multivariate analysis
15
Antibiotic resistance and bacteraemia
(5) Specific pathogens S. aureus Gonzalez et
al, 1999 Madrid, Spain, 1990-1994 S. aureus,
pneumonia bacteraemia prospective cohort study
16
Antibiotic resistance and bacteraemia
(6) Specific pathogen - S. aureus Chang et al,
2003 Multi centre, prospective observational
study of 505 patients in USA. End points were
persistent and relapsed infection Factors relate
to relapse in multi variant analysis -
infective endocarditis vancomycin therapy (vs
nafcillin) for MSSA Outcomes when IE excluded
17
Antibiotic resistance and bacteraemia
(7) Specific pathogen - S. aureus Sakoulas et
al, 2004 30 patients with S. aureus bacteraemia
recruited into clinical trials. (PIII/IV)
treated with vancomycin logistic regression
indicated significant relationship between MIC
(and killing) and treatment success
18
Antibiotic resistance and bacteraemia
(8) Specific pathogen S. aureus Conterno et
al, 1998 Sâo Paulo, Brazil, 1991-92 retrospective
case control study comparing MSSA to MRSA (n
136) Multivariate analysis - 3 risk factors for
death - lung as site of entry OR
17.0 shock OR 8.9 MRSA OR 4.2 MRSA
bacteraemia more likely to have inappropriate
therapy in first 48h
19
Antibiotic resistance in bacteraemia
(9) Specific pathogen P. aeruginosa appropriat
e therapy improves outcome Yes - acute
leukaemia Bodey et al, 1985 Yes - general group
in HIV Vidal et al, 1996 No - general
group Hilf et al, 1989 No - ICU
patients Carmeli et al, 1999 combination
therapy improves outcome Yes - general
group Hilf et al, 1989 Yes - acute
leukaemia Bodey et al, 1985 (monotherapy with
aminoglycoside) No - general group inc HIV Vidal
et al, 1996 No - cancer Chatzinikolaou et al,
2000 (monotherapy with ceftazidime or imipenem)
20
Antibiotic resistance in bacteraemia
(10) Specific pathogen P. aeruginosa Chamot et
al, 2003 115 patients with P. aeruginosa in
historical cohort between 1988-98,
in Switzerland Cox proportional hazard model to
30d follow-up
21
Pseudomonas aeruginosa bacteraemia Zelenitsky et
al, 2003 retrospective study of 38
patients serum concentrations, MIC
determined Outcome measured as - persistent
infection (21) - death to 30d (21) -
cure (58) Cmax/MIC ratio of gt8 predicted gt 90
cure for aminoglycosides and ciprofloxacin
22
Extended spectrum Blactamase (1) Paterson et
al, 1998 400 consecutive blood stream isolates
of K. pneumoniae, 11 hospitals Overall mortality
- 24 Mortality lower if carbapenem used in
first 5 days (5 vs 43, p0.01) 21 mortality
if treated with ciprofloxacin and
susceptible 50 (2/4) mortality with
cefipime 50 (2/4) mortality with
piperacillin-tazobactam combination of active
Blactam plus amikacin did not improve
outcome (mortality 15 vs 17 pgt0.2)
23
Extended spectrum Blactamase (2) Kim et al,
2002 142 blood isolates in Korea, E. coli or K.
pneumoniae Strain MIC gt 2mg/L to 3rd generation
cephlosporins Patients treated with extended
spectrum cephalosporin (most received aminoglycosi
de)
24
Extended spectrum Blactamase (3) Piperacillin -
tazobactam Burgess (2003) ESBL E. coli or
Klebsiella overall 6/18 patients failed 4/9
piperacillin-tazobactam 2/9 other
agents Ambrose et al, 2003 Piperacillin-tazobacta
m 3.375g 6hrly 0.50 - 0.73 target ascertainment
of ESBL positive E. coli, K. pneumoniae in Monte
Carlo simulations (4.5g 8hrly probably similar
4.5g 6hrly better)
25
Extended spectrum Blactamase (4) treatment of E.
coli/Klebsiella with ESBLs in the urinary tract
This page is intentionally left blank
26
Antibiotic resistance in urinary tract
infection Talan et al, 2000 Los Angeles, USA as
part of a randomised double blind comparative
study of ciprofloxacin TMP/SMX conducted
between 1994-7 (n 378) Resistance to TMP/SMX
18 in E. coli (90 of pathogens) TMP/SMX
associated with higher bacteriological/clinical
failures
27
Antibiotic resistance in pneumonia
(1) Definition of penicillin resistance- penici
llin susceptible ? 0.06mg/L intermediate 0.1
- 1.0mg/L resistant ? 2mg/L
28

• Antibiotic resistance in pneumonia (2)
• penicillin non susceptibility does not impact on
  clinical
• response or outcomes for therapy with
  penicillin/amoxicillin
• clavulanate
• paediatric community acquired pneumococcal
  pneumonia (retrospective
• n 207), Friedland Klugman 1992
• adults with pneumococcal pneumonia
  (retrospective n 23)
• Sandches et al 1992
• paediatric bacteraemic pneumococcal infection
  (prospective), Friedland,
• 1995
• adults with pneumococcal pneumonia
  (prospective n 504) Pallarres
• et al, 1995
• invasive pneumococcal infection bacteraemia
  (retrospective n 106)
• Choi Lee, 1998

29

• Penicillin non susceptibility does not impact
  (continued)
• paediatric invasive pneumococcal infection,
  mainly bacteraemia
• (retrospective) Deeks et al, 1999
• hospitalised patients with pneumococcal
  community acquired
• pneumonia (retrospective n 101 pen R ?
  2mg/L) Ewig et al, 1999
• hospitalised patients with pneumococcal
  bacteraemia (retrospective
• n 156) Farinas-Alvarez et al, 2000
• community acquired pneumococcal pneumonia
  (prospective, n 465)
• Bedos et al, 2001
• hospitalised patients with invasive
  pneumococcal pneumonia
• (prospective, n 146) Moroney et al, 2001

30

• Penicillin non susceptibility does have a
  clinical impact
• pneumococcal pneumonia (n 5837)
• overall mortality related to older age
• underlying disease
• Asian race
• living in Toronto
• Excluding early deaths i.e. lt4 days-

Feikin et al, 2000
31

• Penicillin non susceptibility does have a
  clinical impact
• pneumococcal pneumonia (retrospective study, n
  462)
• multivariate analysis identified the following
  as independent
• predictors of mortality - older age
• severe disease
• multilobar infiltrate
• effusion on CXR
• hispanic
• high level penicillin resistance
• Turrett et al, 1999

32

• Penicillin non susceptibility does have a
  clinical impact
• adults with bacteraemic pneumococcal pneumonia
  (n 192)
• gt increased risk of suppurative complication
  after adjustment
• for other factors
• Metlay et al, 2000
• children with invasive infection -
• mainly bacteraemia (n 304)
• gt longer ITU stay all other factors similar
• Quach et al, 2000

33
Conclusion for S. pneumoniae- penicillin
resistance probably only has therapeutic signifi
cance once MIC values are ? 2-4mg/L
34
Putting it together (1)
microbiological cut offs
wild type distributions
MIC
pD index pharmacokinetics microbiological
outcomes clinical breakpoints clinical
outcomes
35
Putting it together (2)
microbiological cut off
Wild type distributions
Penicillin-S. pneumoniae
MIC
most drug-bacteria cephalosporin-ESBL Blactam-MRSA
pD index pharmacokinetics microbiological
outcome clinical outcome
Vancomycin-MSSA P. aeruginosa P/T - ESBLs
36
Conclusions gt clinical data on resistance
significance is weaker than animal/in vitro
data gt appropriate early therapy probably
improves patient outcomes gt applies in a wide
range of clinical contexts and pathogens but
not everywhere gt categorical sensitivity testing
(S/I/R) is a crude approximate of the true
drug - pathogen - host relationship gt clinical
breakpoints should have improved predictive
value as pD principles are understood gt
microbiological breakpoints may be
therapeutically misleading