# High and strikingly early failure-rate following gram-negative periprosthetic joint infection – a retrospective cohort study on 72 cases

## TL;DR

Gram-negative joint infections after surgery have high early failure rates, with some bacteria like Pseudomonas leading to worse outcomes.

## Contribution

This study provides new insights into the poor outcomes of gram-negative periprosthetic joint infections and evaluates treatment strategies.

## Key findings

- Only 43.1% of gram-negative PJI cases were infection-free and revision-free.
- Pseudomonas aeruginosa had the lowest success rate at 18.2%.
- Two-stage revision showed better success than single- or multi-stage revisions.

## Abstract

Periprosthetic joint infections (PJI) represent a major complication of total joint arthroplasty. While most infections are caused by staphylococci species, a notable proportion involves gram-negative bacteria. Due to the smaller numbers, outcome reports in literature are scarce and heterogenous success rates have been reported. This study aimed to (1) evaluate the overall treatment success of gram-negative PJI and (2) identify the most suitable surgical treatment strategy in eradicating gram-negative PJI.

Seventy-two cases of gram-negative PJI treated between 2010 and 2022 were analyzed in this retrospective cohort study. The median follow-up (IQR) was 18.0 (46.0) and at least 12 months. Outcomes were assessed based on the 2013 Delphi consensus on PJI outcome. PJI was defined according to the EBJIS-criteria.

The overall infection-free and revision-free survival rate was 43.1% (31/72) and 56.9% (41/72). 32 out of 41 treatment failures (78.0%) appeared within the first 3 months. Among the causative pathogens, Pseudomonas aeruginosa-related PJI had the poorest outcome resulting in an infection-free survival of only 18.2% (2/11), whereas infections caused by Enterobacter cloacae had the highest success rate of 58.3% (7/12); (p = 0.12). Success rates were 65.0% (13/20) for two-stage revision, 38.1% (8/21) for multi-stage revision, 36.0% (9/25) for DAIR and 16.7% (1/6) for single-stage revision. In total, 6 amputations, 2 knee arthrodesis, and 8 resection arthroplasties were required for definitive infection eradication.

Gram-negative PJI may follow a more aggressive course with earlier and higher failure rates than previously thought. DAIR may be an option in selected cases, while two-stage revision showed a trend towards a more favorable infection-free survival in comparison to single- and multi-stage revisions. The results suggest pathogen-specific differences may influence outcomes and support individualized treatment strategies, warranting confirmation in large, prospective, multicenter studies.