Despite the excellent long-term results reported about total hip replacements regarding joint osteoarthritis (Synder et al., 2012; Chang and Haddad, 2020; Lucchini et al., 2021), orthopedic surgeons continue to struggle against a devastating complication: periprosthetic joint infection (PJI) (Kapadia et al., 2015). In primary arthroplasties, their incidence is about 0.5–3 %, but in hip revision, this number increases to 14.8 %. (Phillips et al., 2006; Leone et al., 2010; Parvizi et al., 2010).

Several authors consider two-stage revision to be the gold standard for chronic PJI, with success rates exceeding 90 % (Amanatullah et al., 2018; Pignatti et al., 2010; Younger et al., 1998, 1997). This strategy involves the removal of all prosthesis components, surgical debridement, irrigation, and implantation of an antibiotic-impregnated cement spacer. (Hsieh et al., 2004). The benefit focuses on occupying a “dead space” and being able to release high local concentrations of antibiotics into the joint environment (Grassi et al., 2007).

There are several types of hip spacer designs, and according to their manufacture we can mention those handcrafted in the operating room and those commercially available (Veltman et al., 2016). Among the advantages of the latter spacers is that their use shortens surgical times, and both types allow for maintaining joint mobility, the absence of soft tissue contraction, and ensures easier reimplantation (Konstantinos et al., 2006).

Since the 2000s, prefabricated antibiotic-impregnated cement spacers (Subiton^®; Laboratorios SL, BA, Argentina) have been used in our field for the treatment of chronic PJI, and to our knowledge, there are no studies about their outcomes that compare when treating infections with an isolated and with multiple microorganisms.

The primary objective of this study was to assess the efficacy of a two-stage revision approach using a commercial prefabricated antibiotic-impregnated cement hip spacer for the treatment of hip PJI regarding monomicrobial and polymicrobial infections. Secondly, to evaluate risk factors for failure of the two-stage revision.

With the Institutional Review Board`s approval, we conducted a retrospective study of patients undergoing revision of THA between January 2002 and January 2018.

Inclusion criteria for the study were adult patients (>18 years) with a diagnosis of chronic hip PJI (>4 weeks of primary THA) who underwent a two-stage revision using a prefabricated gentamicin-impregnated cement of polymethylmethacrylate (PMMA) hip spacer (Subiton^®, Laboratorios SL, Buenos Aires, Argentina) and completed a minimum follow-up of 3 years after the second stage.

Patients with previous hip revision, infections with negative culture, and those who did not complete the two stages of revision were excluded from the study.

PJI was defined according to the 2018 Musculoskeletal Infection Society (MSIS) criteria (Parvizi and Gehrke, 2018). Patient medical records were reviewed, and the following information was registered: gender, age, body mass index (BMI), Charlson comorbidity index (CCI), smoking (patients that were smokers at the time of the first stage), involved microorganisms, length of hospital stay, antibiotic therapy duration, and follow-up. We also documented serum levels of erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and white blood cell (WBC) count before the first and second stages.

In the first stage approach, there was a period of 10–14 d in which broad-spectrum intravenous antibiotic therapy was administered postoperatively. After this, it was continued with selective oral therapy that was adjusted according to the resistance profile of the infecting microorganism, guided by the infectious disease department.

A normal distribution test was performed. Continuous variables were described as mean and standard deviation or median and interquartile range (according to normality) and categorical variables as frequency and percentage. Comparison of clinical and functional outcomes were assessed with the Student's t test and categorical variables were analyzed with the chi-square (X2) test (or Fisher's exact test if needed).

To calculate risk factors, continuous variables were transformed into categorical ones as follows: age – <70 and >70 years old, and BMI – 20–24, 24–30, and >30 kg m-2. After this, univariate logistic regression was calculated to identify potential risk factors for failure. A difference of p<0.05 was considered statistically significant. All the data were collected into an Excel^® (Microsoft, Redmon, USA) spreadsheet, and statistical calculations were performed with the software GraphPad Prism 8.0^® (LaJoya, CA, USA).

Prefabricated antibiotic PMMA hip spacers have initially been impregnated with aminoglycosides, but recent designs have allowed the addition of vancomycin. Although they traditionally contain low doses of antibiotic, they can achieve higher local concentrations because of their dimpled surface (Scott, 2020).

The main finding of our study is that we observed that the use of prefabricated antibiotic-impregnated cement spacers is an effective tool for the treatment of PJI, with a success rate of 90.5 % and with no significant differences regarding functional outcomes between patients with monomicrobial or polymicrobial infections.

Although we have not found previous series reporting the use of the spacers used in the present study, comparatively, our success rate was as reported by Hsieh et al. (2004). These authors assessed 40 patients with a recurrence rate of 2.5 % with 4 years of follow-up. Likewise, Romanó et al. (2011) in a prospective cohort study of 20 patients who underwent a two-stage hip revision using antibiotic-impregnated articulating hip spacers reported a 95 % success rate.

Polymicrobial hip PJI has been traditionally considered a risk factor for failure in hip revision (Della Valle, 2011), and its incidence is between 6 and 37 % (Pulido et al., 2008; Holleyman et al., 2016; Moran et al., 2007; Peel et al., 2012). In our series, both cohorts achieved promising success at the end of the study, and though we observed a higher percentage of failure (12.5 %) in the polymicrobial group, the difference versus patients with monomicrobial (8.3 %) joint infections was not statistically significant. These outcomes agree with the findings of Bozhkova et al. (2016): they reported a higher failure rate in patients with polymicrobial infections (34.1 %) compared to the monomicrobial cohort (23.4 %) after two-stage revision, but this was not statistically significant. Another finding is that regardless of whether it was monomicrobial or polymicrobial PJI, both achieved statistical improvements after second-stage surgery regarding WOMAC and mHHS scores, with no significant differences between them.

Previously, other studies have described similar findings: Pignatti et al. (2010) reported values greater than 80 after 5.3 years of follow-up, analyzing 60 patients that underwent a two-stage hip revision for PJI treatment.

In addition, significantly higher levels of CRP and WBC were observed in the polymicrobial group when comparing monomicrobial infections after the second stage, but within normal ranges. Bozhkova et al. (2016) also reported higher percentages of CRP levels in patients with polymicrobial (42.6 %) joint infections with failure of second stage, though this was not significant. Similarly, Mortazavi et al. (2011) analyzed 117 patients that underwent two-stage revision for PJI from a prospective database and failed to find an association between CRP and ESR levels with failure. The authors of this study agree that there are no clear cut-off values of these biomarkers to predict failure. We consider every patient as a singular case, searching for a combination of laboratory, radiographic, and clinical parameters to decide on reimplantation.

Logroscino et al. (2019) reported that patients with BMI >25 had significant association with reinfection after two-stage hip revision. Likewise, Jhan et al. (2017) assessed 61 cases of hip PJI that underwent two-stage revision and observed that BMI >30 kg m-2 was associated with higher rates of failure.

In concordance with these authors, uni- and multivariate analysis showed a statistically significant association between failure and BMI >30 kg m-2 in our study.

Furthermore, this study found that smoking patients were 12.94 times (CI95 % 3.17–60.41) more likely to fail after reimplantation than non-smokers. This agrees with Ahmad et al. (2019), who found a 3.9 (CI95 %1.1–14.6i) increased risk of failure in smokers after analyzing 97 patients undergoing two-stage revision for PJI.

In addition, the group of Parvizi (Aali Rezaie et al., 2018) described a significant association between greater CCI and failure after reimplantation in regression analysis with an odds ratio of 1.40 (CI95 % 1.06–1.86). Although our findings do not agree with the latter, we must be careful, because there is a trend to believe that the greater the comorbidities of the patients, the greater the rate of failure.

Our study is not without limitations: we must mention that it is a retrospective study, and we did not consider every possible preoperative comorbidity, which might represent a confounding factor. On the other hand, this paper analyzed an adequate number of patients that were operated on in the same institution by the same trained hip-revision surgeons. We have the task left to continue with long-term follow-up and to establish a control group to give greater relevance to our findings.

Our study suggests that gentamicin-impregnated PMMA hip spacer is an effective tool for the treatment of PJI, achieving similar outcomes whether it is monomicrobial or polymicrobial infections. Randomized prospective studies are needed to obtain more reliable conclusions.