Efficacy of colistin-loaded cement spacer in carbapenem-resistant Klebsiella pneumoniae experimental prosthetic joint infection

Background:

Parenteral administration of colistin carries a high risk of adverse effects, including nephrotoxicity and neurotoxicity. Local use of colistin may avoid systemic effects, while ensuring adequate concentrations in situ. In a new rabbit model of carbapenem-resistant Klebsiella pneumoniae (CRKP) knee-prosthesis infection, we evaluated the efficacy of colistin-loaded cement spacer, alone or in combination with intramuscular (i.m.) colistin.

Methods:

A partial knee replacement was performed with a silicone implant. Seven days after infection by KPC-99YC (colistin MIC 0.5 mg/L), surgical debridement with removal of the infected prosthesis was performed, and rabbits were randomly assigned to one of 4 treatment groups of 12 rabbits: prosthesis replacement by drug-free cement spacer (control); prosthesis replacement by colistin-loaded cement spacer (3 MUI of colistin/40 g of cement); prosthesis replacement by drug-free cement spacer and i.m. colistin (150 000 IU/kg im tid for 7 days, equivalent to 3 M IU tid in humans); or colistin local + i.m.

Results:

Colistin local or i.m did not increase the proportion of rabbits with sterile bones as compared to controls. However, combination of local and i.m. colistin was more effective than controls, both in terms of sterilization rate and bacterial load.

Treatment	No. of rabbits with sterile bone/total no. of rabbits	P versus control group (significant if <0.05)	log10 CFU/g of bone (mean ± SD)	P versus control group (significant if <0.05)
Control	0/10		5.9 ± 1.6
Colistin local	3/12	0.590	4.0 ± 2.4	0.206
Colistin i.m.	3/10	0.311	4.1 ± 2.0	0.043
Colistin local + i.m.	5/9	0.049	2.5 ± 1.5	0.005

Conclusion:

Combination of systemic and local colistin is a promising option to optimize the treatment of carbapenem-resistant Klebsiella pneumoniae prosthetic joint infection.