Hospital-acquired Pneumonia and Ventilator-associated Pneumonia

Recent Advances in Epidemiology and Management

François Barbier; Antoine Andremont; Michel Wolff; Lila Bouadma


Curr Opin Pulm Med. 2013;19(3):216-228. 

In This Article

Antimicrobial Treatment of Hospital-acquired Pneumonia and Ventilator-associated Pneumonia

Delayed effective therapy increases hospital mortality in HAP/VAP patients, making the choice of empirical drugs a crucial dilemma.[1,6] To improve the likelihood of adequate coverage, current guidelines ( Table 3 ) usually recommend an antipseudomonal combination therapy except for early-onset pneumonia with no risk factor for MDR bacteria.[1,8–10] Facing the epidemic of ESBL-producing enterobacteria and other MDR-GNB, antipseudomonal carbapenems (imipenem/cilastatin and meropenem) have become the most empirically prescribed β-lactams in European ICU for HAP/VAP.[51] Despite limited lung diffusion,[10] aminoglycosides should be preferred to fluoroquinolones, given the resistance frequencies in P. aeruginosa and the possibility to achieve bactericidal lung concentrations with a high-dosing regimen (e.g. 25 mg/kg for amikacin). Empirical indications of polymyxins in patients at risk for carbapenem-resistant GNB should be clarified in updated guidelines.[107]

Re-assessment is mandatory at day 2–3, both to correct inadequate regimens when appropriate and to avoid an overconsumption of broad-spectrum drugs.[1,8–10] Antimicrobials must be stopped when the diagnosis is not bacteriologically validated, except for particular situations.[10] When pneumonia is confirmed, the antimicrobial spectrum must be narrowed whenever possible on the basis of susceptibility testing. Monotherapy is conceivable provided that the empirical regimen was adequate, the patient's condition improves and the causative pathogens do not exhibit extensive resistance patterns.[10] Even not validated in HAP/VAP, retrocession to β-lactam/β-lactamase inhibitor might be discussed for ESBL-producing enterobacteria with minimal inhibitory concentration (MIC) below the break-point values.[108,109] A control sample should be obtained to assess bacterial load kinetics and detect precocious resistant mutants.[10] The 8-day standard duration of treatment can be safely shortened by monitoring plasma PCT levels.[100] Longer treatments may be discussed in cases of immunosuppression, unfavourable clinical response or extensively drug-resistant pathogens.[10]

Colistin appears as effective as other antimicrobial classes for VAP caused by MDR-GNB.[110] The optimal dosing regimen of this concentration-dependent antibiotic is still debated; however, an intravenous loading dose of 9 million units followed by 4.5 million units twice daily (with protocolized adaptation in case of renal failure) may be an adequate scheme,[111] in accordance with previous pharmacokinetic/pharmacodynamic data,[112] the median colistin dose being correlated with clinical and microbiological success rates.[113] The incidence of colistin-induced nephrotoxicity is 18–43%, especially with a high-dosing regimen or when other nephrotoxic drugs are prescribed.[111,113,114] The nephroprotective effect of ascorbic acid co-administration during colistin therapy should be evaluated in clinical practice.[115] Experimental data suggest that inhaled colistin should be used as an adjunctive therapy to reach high lung concentrations,[116] but available clinical evaluations yielded ambiguous results,[117–120] and further RCTs are needed to clarify this issue. Other inhaled antimicrobials have been tested in experimental or clinical VAP caused by MDR-GNB, namely ceftazidime,[121,122] imipenem/cilastatin,[123] amikacin[124] and tobramycin.[118,125] Nebulized ceftazidime/amikacin combination does not improve the course of clinical P. aeruginosa VAP when compared with intravenous administration.[122]

Both vancomycin and linezolid are recommended for the therapy of suspected or proven MRSA pneumonia.[1,8–10] In experimental models, linezolid is associated with faster bacterial clearances and a lower histological severity of pneumonia,[126–128] with better pharmacokinetic/pharmacodynamic profiles[127] and a possible immunomodulatory effect.[129] However, previous meta-analysis did not demonstrate a superiority of linezolid in terms of clinical cure and bacteriological eradication.[130,131] In a recent RCT,[132] linezolid was associated with a higher clinical success rate and a lower incidence of renal impairment than vancomycin, but all-cause 60-day fatality rates were similar in both groups. We believe that linezolid should be preferred to vancomycin when other risk factors of acute kidney injury are present and for HAP/VAP caused by MRSA strains with vancomycin MIC higher than 1 mg/l, a cut-off associated with vancomycin therapy failure.[133,134]

New antibiotics that may complete the therapeutic arsenal for HAP/VAP are scarce. Lipoglycopeptides (i.e. dalbavancin, oritavancin and telavancin) are bactericidal for MRSA strains with high vancomycin MIC.[135] Telavancin is not inferior to vancomycin for the treatment of MRSA pneumonia, but may be associated with an increased incidence of acute kidney injury.[136] Ceftaroline[137,138] and ceftobiprole,[139] two new-generation cephalosporins, and iclaprim,[140] a dihydrofolate reductase inhibitor, are also active against MRSA. None of these drugs are currently approved by the US Food and Drug Administration (FDA) for the treatment of MRSA HAP/VAP. Novel carbapenem-sparing molecules for the treatment of MDR-GNB include temocillin,[141] a 6-hydroxymethyl-ticarcilline with activity against ESBL-producing and AmpC-producing enterobacteria, and avibactam (formerly NXL104),[142–145] which restores the activity of various β-lactams (e.g. ceftazidime) in isolates producing Ambler class A (ESBL and KPC-type carbapenemase) and AmpC β-lactamases (including P. aeruginosa); these two promising antibiotics remain to be evaluated in HAP/VAP.