Pharmacokinetic/pharmacodynamic modelling of polymyxinB, rifampicin and tigecycline against pandrug-resistant Acinetobacter baumannii in an invitro model

Abstract number: O517

Lim T.P., Lee W., Sasikala S., Tan T.Y., Hsu L.Y., Tan T.T., Kwa A.L.

Objective: Outbreaks of pandrug-resistant (PDR) Acinetobacter baumannii (AB) have emerged in Singapore. Combination therapy is often the remaining viable option until new antibiotics are available. While polymyxin B (P) may remain a viable treatment option, heteroresistance has become a major problem. We evaluate the efficacy of P, rifampicin (R) and tigecycline (T) combined against PDR AB isolated from our local hospitals.

Methods: PDR AB isolates from all public hospitals in Singapore were collected from 2006–07. MICs were determined according to a modified CLSI broth-dilution method. Time-kill studies (TKS) were then performed with the maximum, clinically achievable, unbound concentration (mg/L) of P (2), R (2) and T (2) alone and in combination against the PDR AB isolates. A hollow-fiber infection model (HFIM) was used to validate our quantitative assessment of combined killing against 2 isolates (selected based on the unique genotype that represents the PDR AB population). Resistance selection of the 2 isolates against P alone in the HFIM were quantified using drug-free and selective (P at 3x MIC) media.

Results: Among 361 non-repeat AB isolates screened, 29 PDR AB isolates found were susceptible to P (MICs 1–2 mg/L) and resistant to all antibiotic classes whereas R MICs ranged from 2–16 mg/L. In TKS, P, R and T alone was bacteriostatic with regrowth by 24 h in all isolates. P+R, P+T and R+T achieved >99% kill from baseline in 15/29, 14/29 and 8/29 isolates with no regrowth at 24 h. These assessments were consistent with observation in HFIM studies where we observe bacterial killing up to 120 h with P+R. Pharmacokinetic validation of the HFIM studies were satisfactory. Minimal killing of the 2 isolates was seen when exposed to P alone. Regrowth was seen at 24 h due to selective amplication of resistant sub-population(s) on P supplemented plates. Repeat MIC testing of the resistant isolates confirms P resistance (MICs 32–128 mg/L).

Conclusions: We have shown that our PDR AB has the propensity to exhibit heteroresistance and combination therapy with P is needed. P+R may be a potential antibiotic combination as therapy for PDR AB infections. The in vivo relevance of our results warrants further investigations.