Introduction

Periprosthetic joint infection (PJI) that develops after total hip arthroplasty (THA) continues to be a substantial economic and physical burden for patients and the healthcare system [1–4]. The reported incidence of PJI is from 0.2 % to 3 % following primary THR, with the average rate at about 1.6 %. However, the absolute number of infected cases has been increasing as the total number of THRs constantly grows.

Despite that all possible measures are undertaken to avoid re-infection after revision of infected THA, its rates remain rather high and range between four and 33 % [4–7]. Each revision results in a greater bone tissue loss and a shorter period of implant survival. Furthermore, implant removal is an invasive operation. Patients’ psycho-emotional suffering and fear of re-infection are also significant factors that should be considered. Therefore, much effort has been undertaken to reduce the burden of infected THA.

Numerous studies have been already published on the outcomes of revision for infected hip and knee arthroplasty [8–12]. Harvesting material for microbiological study is a very significant part of any THA revision operation that aids to identify infecting micro-organisms and plan antibiotic strategies.

Our study aimed to delineate and analyze the infecting micro-organisms at our bone infection unit pre-operatively and intra-operatively at the first-time revision THR due to chronic infection, early and late re-infection micro-organisms, with the objective to reveal the bacterial spectrum to fight the first-time infected THA and re-infected THA.

Materials and methods

Seventy-three patients were admitted to our bone infection unit for revision due to chronic PJI between 2004 and 2014. There were 46 males and 27 females in the mean age of 51.93 ±10.9 years. Clinical signs of infection were sinuses (89 % of cases), wounds (8 %), hyperemia and swelling (3 %), local rise of temperature, pain or joint function disorders. Infection location was assessed by fistulography findings. At the moment of the revision operation, the infection manifestation time was more than a month in all the cases. Debridement protocol with implant retention was not possible to perform. Revision required debridement and removal of implant components in all cases.

Forty-three patients had a two-stage revision THA with the use of cemented spacers impregnated with antibiotics at stage 1. Thirty patients were treated with a modified resection arthroplasty (MRA) using the Ilizarov apparatus [13]. The indications to MRA were bone defects of grade II to IV (Paprosky), numerous previous operations, soft-tissue deficit, and severe immune deficiency. Several patents of this group rejected revision THA.

In both groups, antibiotics were administered according to susceptibility tests in the maximum dosage starting from the operation day with intravenous infusion for two to three weeks followed by oral antibiotics prescribed for three weeks. Anti-thrombosis preparations were administered for 35 days.

Microbiological study was conducted in all of them. The findings were studied retrospectively.

Pre-operatively, the object for microbiological study was the discharge from wounds and sinuses. The joint was punctured in case of absent pre-operative sinuses and wounds.

Intra-operatively, the material was collected from the soft tissues prior to approach to the joint. Joint fluid was then collected as well as the material from all the implant component surfaces and cements (five to six tissue samples per patient). Once surgical debridement was completed, the tissues samples from the areas debrided were sent for control in order to evaluate the efficiency of debridement (two to three control tests). Similar microbiological diagnostic tests were performed in all the cases of re-infection.

Identification of microbe genera and species was performed with classical methods (study of their tinctorial, culture, and biochemical properties) as well as with the use of bacterial analyzers (ATB Expression, Bio Merieux, France; Walk Away-40, USA) supplied with microtests and microbiological laboratory software (WHONET 5.6). Sensitivity to antibiotics and quantitative evaluation of microbes in the tissue samples was assessed.

Patients were also examined on the presence of risk factors that may aggravate the condition such as the number of previous surgery on the hip, diabetes, as well as femoral and acetabular defects.

Mean follow-up after treatment was 4.30 ±1.77 years (range, 2–11 years).

Results

Only 37 patients (50.7 %) had a complete correspondence of the pre-operative microbial species to the ones detected in the material taken intra-operatively. According to intra-operative tests, gram-positive single genera were identified in 35 patients (48 %). Staphylococcus species were most common single germs detected in 97 % of all single genus infections. Staphylococcus aureus prevailed (74 %). Methicillin-resistent staphylococcus aureus (MRSA) and epidermal staphylococcus (MRSE) were 17 %. Microbe associations (Table 1) were present in 33 patients (45 %). They were two-component in 23 patients (31.5 %). Three-genus infection was detected in ten cases (13.5 %). Staphylococcus species were found in 57 % of mixed infection cases. One patient did not have any bacterial growth. Four cases (5.5 %) of single gram-negative microorganisms were Acinetobacter baumannii, Burkholderia cepacia, Enterobacter sp., Serratia marcescens.

Table 1 Microbes in associations detected in patients with chronic periprosthetic infection
Full size table

Control tests after debridement did not show any microbe growth. However, re-infection developed in a total of 29 patients (25 early and nine late recurrences) (Table 2). Five out 29 patients had both early and late re-infection. Twenty-four cases out of those 29 were affected with staphylococcus species at the index surgery. Staphylococcus infection was identified in 19 out of 24. The same germs of staphylococcus repeated in 16 cases. Microbial associations were identified in 13 patients. Gram-negative bacteria repeated or joined in 11 of them. Acinetobacter sp. and P. aeruginosa isolates were most common. Complete correspondence of laboratory findings with the ones at index revision surgery was identified in 13 patients with re-infection. All but one were affected with staphylococcus species. S. aureus prevailed. Only one patient with single genus gram-negative infection had an early recurrence.

Table 2 Reinfection cases and their microbes
Full size table

The findings on the risk factors that could have a negative effect on eradication of infection are presented in Table 3.

Table 3 Frequency of risk factors in the patients with infection recurrence and without it
Full size table

In the MRA group, all re-infected cases were debrided with the Ilizarov apparatus on and continued fixation. MRA yielded 93.3 % of final success in fighting infection (28/30 patients) while in the THA group it was 83.8 % (36/43). Two patients with late reinfection were not revised due to associated conditions.

Discussion

It is accepted that a thorough history, physical examination, complete set of radiographs and appropriate laboratory tests including C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), synovial fluid leukocyte and neutrophil counts as well as joint aspiration and tissue biopsy are essential in the initial evaluation of suspected infected THA [14, 15]. Aspiration and nuclear imaging are helpful in doubtful cases. ESR or CRP is performed in all patients with a suspected PJI when the diagnosis is not clinically evident. The combination of an abnormal ESR and CRP seems to provide enough evidence in suspected chronic PJI along with clinical signs. Although a number of diagnostic criteria have been proposed by the latest consensus, a gold standard for PJI diagnosis is still lacking [16, 17].

A meticulous study of PJI agents in the material harvested for laboratory culture tests was able to detect the bacteria and the contamination grade at all the stages in our series. Collection of such a material is a very important part of revision in infected THA. However, only intra-operative tests revealed a true spectrum of germs that should be targeted by antibiotic therapy.

As reported, gram-negative bacteria caused infection in 7 % of THA cases, which is similar to our series at index revision [9, 10]. We had one early recurrence in this group but the species could not be detected as the patient received antibiotics prior to second debridement performed early after the index surgery [15, 18]. However, we observed high incidence of gram-negative bacteria in microbial associations in re-infection cases.

It was confirmed by numerous studies that gram-positive bacteria are the main agents that are responsible for 75–88 % of infected cases following primary THA [8, 11, 12, 14, 18]. Gram-positive bacteria were the main causative micro-organisms in our patients with chronic PJI that were either single genera or mixed with other microbes. Our data confirmed the findings of other studies that the staphylococci were the most common species responsible for primary infection and re-infection [11, 12, 18].

Several studies showed low re-infection rates when a two-stage THA revision was used [19–22], but the most recent studies in the field have shown that even a two-stage THA may not be so successful due to associated medical comorbidities [23–25]. Despite the fact that the implants were removed in our series, the success rate in the two-stage revision THA in fighting infection corresponded to the average reported in the latest literature [22–24].

Current antibacterial preparations are able to fight gram-positive bacteria, including the resistant species such as S. aureus (MRSA) and S. epidermidis (MRSE). However, the recurrence rate in our series was rather high due to several reasons that provoked re-infection. First, the re-infection cases were mainly caused by resistant strains or associations of microbes [8, 12]. It is also possible that bone debridement failed to remove all the infected tissues and the bacteria could be present as biofilm remnants in the para-articular tissues. Other reasons that should be considered are possible infection of a post-surgical haematoma, previous surgery on the hip, deficit of soft tissues, post-traumatic OA, polymicrobic and resistant infection, prolonged operation time, prolonged wound drainage, poor patient’s immune state or age [1, 8, 25]. Therefore, we studied the factors that have a negative effect in fighting infection (Table 1). The combination of these factors was clearly seen in the cases of infection recurrence.

The high rate of acute reinfection in the MRA group might have been due to the haematoma in the residual bone marrow canal of the femur and application of wires and half-pins of the Ilizarov apparatus that was an additional trauma. The main canal of microbe penetration with the use of this technology is wire- and half-pin tracts. Antibiotic therapy and debridement repeated in them within a month resulted only in two late re-infections that were treated with pin-tract sanation measures and antibiotics only. On the contrary, there were seven late re-infections in the THA group. However, our study was not aimed at the comparison of the treatment method groups. We focused on the impacts of infection species and associated factors [8].

Multiple intra-operative bacteriological tests provided a true picture of infecting agents and their sensitivity to design individual protocols of antibiotic therapy. Resistant strains of micro-organisms and S. aureus were the main targeted microbes by administration of antibiotics. Nowadays, the most effective antibiotics could be combined in fighting aggressive micro-organisms but there is no uniform approach as to whether to use them alone or combined [14, 18]. Disputes have been held in regard to the duration of antibiotic administration. The standard period of intravenous antibacterial therapy according to susceptibility tests in our hospital is two to three weeks followed by oral administration for three to four weeks. The primary requirement is to achieve maximum high concentrations of effective antibiotics in the bone and para-articular soft tissues that would be able to eradicate the pathogenic micro-organisms [17].

When microbial species were compared, it was obvious that the pre-operative findings on the micro-organisms coincided only 50.7 % with the ones that were obtained after the intra-operative material had been studied. This confirms the findings of several recent studies [26, 27]. Such a disparity in the spectrum of micro-organisms detected could be explained by the fact that the true infection focus is located in the deepness of para-articular structures. Therefore, the infection agents and grade of contamination detected intra-operatively at the first-time surgery for infected THA present reliable findings for assessment of microbial resistance to antibiotics, development of antibacterial treatment protocols, and for prognostic purposes.

There are few studies that reported on the re-infection microorganisms after the revision of infected THA [28, 29]. The bacterial spectrum of re-infection in our series showed that both prophylactic measures at first-time revision for infected THA and therapeutic antibiotic regimes due to re-infection should be focused on targeting staphylococci. However, Acinetobacter sp. and P. aeruginosa isolates were frequent species in reinfection cases [29]. Therefore, diagnosis, antimicrobial susceptibility testing, and antibacterial treatment along with a proper choice of a surgical protocol are the most significant issues in the arrest of infection [14, 17].

Conclusion

A thorough study of periprosthetic infection organisms in the material harvested intra-operatively is most essential in revision of infected THA. Prophylactic and therapeutic antibiotic regimes should be focused on targeting staphylococci.