Debridement, Antibiotics and Implant Retention (DAIR) Has High Success Rate for Acute Periprosthetic Joint Infection

Background: The efficacy of debridement, antibiotics and implant retention (DAIR) for acute periprosthetic joint infection (PJI) has been debated. The purpose of this study was to report our clinical outcome and experience of managing acute PJI with DAIR. Methods: A retrospective review of all patients diagnosed with acute PJI after hip or knee replacement surgery and managed with DAIR in our prospective joint replacement register from 2008 to 2019 was performed. Patients’ symptom onset duration, inflammatory marker levels, bacteriology, and surgical outcome were tracked and recorded. Results: A total of 24 patients with 7 after hip replacement and 17 after knee replacement were included. Twenty-one were early post-operative infection and 3 were late acute haematogenous infection. During a mean follow-up time of 29.2±15.1 months, twenty-two were successfully treated while 2 failed who required repeated DAIR. The overall success rate of DAIR was 91.7%. For staphylococcal infection, DAIR has 100% success rate. Five patients presented with symptoms between 4 to 8 weeks also achieved 100% success rate. Conclusions: DAIR has a high success rate for acute PJI, and can be performed in selected patients whose symptoms have sustained for over 4 weeks. DAIR has high rate of success for staphylococcal infection. Metagenomic next-generation sequencing (mNGS) test can be used as a potential tool to identify pathogens in acute PJI.

Periprosthetic joint infection (PJI) after total hip and total knee arthroplasty is a rare but catastrophic complication that can cause significant morbidity to the patient and cost to the health care system [1]. It was found from the Medicare inpatient data set that the 1- year and 5-year risk of PJI was 0.69% and 1.09% for THA and 0.74% and 1.38% for TKA, respectively [2]. Though not commonly seen, treating PJI has already been estimated to cost the USA $566 million in 2009, and the number is projected to reach $1.62 billion in 2020 [1].
There have been several risk factors proposed to affect the outcome of DAIR, including age, American Society of Anesthesiologists (ASA) score, diabetes mellitus (DM), rheumatoid arthritis, and steroid therapy [18]. On the other hand, the role of optimal timing of DAIR, antibiotic regime, the involved microorganism, and whether or not to exchange the polyliner are still unclear. Therefore the purpose of this study is to report our outcome and experience of managing acute PJI with DAIR.

Methods
A retrospective review of our prospective joint replacement registry from 2008 to 2019 was carried out. Patients who were diagnosed with acute PJI after hip or knee replacement surgery and managed with DAIR were included. The diagnosis of PJI was based on the 2011 Musculoskeletal Infection Society (MSIS) PJI diagnostic criteria [19]. Both the early postoperative infections (Tsukayama type 2) and the late haematogenous infections (Tsukayama type 3) were included [3]. 'Acute' was defined as infections happening within 3 months after the primary surgery, or symptom onset within 3 months in a late wellfunctioning joint, as was proposed by Zimmerli [20].
Our center was a tertiary referral center in Fujian province, which has a population of 38 Culture has also been a standardized procedure in our center. Joint fluid, once collected, was injected into aerobic, anaerobic bottles, and inoculated to fungal plate. A small amount of fluid was also injected into pediatric blood culture bottle. For tissues, they were collected by blade cut instead of electrode, and further cut into pieces, stored into sterile container and sent for culture immediately from operation theater to the department of microbiology by designated personnel. Culture was routinely performed for 7 days, but in case of negative results or suspected low-virulence pathogen, culture was prolonged to 14 days. A paired sample of joint fluid and tissues were also sent for mNGS exam, which was carried out by the BGI group company (Shenzhen, China). The results were normally available in 48 hours.
An empirical antibiotics regime of vancomycin combined with ceftazidime were prescribed until culture/mNGS results were available. It was then changed to pathogen-specific antibiotics according to the drug susceptibility results. In general circumstances, intravenous (IV) antibiotics were used for 2 weeks followed by oral antibiotics for additional 4 weeks.
Surgical outcomes were defined as successful if patients' clinical symptoms had been relieved, and inflammatory marker levels including CRP, ESR and WBC count had returned to normal, X-rays showed no prosthetic loosening, and no lifelong antibiotics suppression was required, at a minimum of one-year follow-up. Outcomes were defined as failed if patients required any further surgeries (e.g., additional DAIR or any forms of debridement, or one or two-stage revision), or needed lifelong antibiotics suppression.
All the patients' baseline characteristics were documented, including age, primary diagnosis, time interval of primary TKA to symptoms, time lag between symptoms onset and debridement, pre-operative CRP, ESR, WBC, lymphocyte, bacteriology, and synovial total cell count. The blood samples for inflammatory markers including CRP, ESR, and WBC were specified as being taken within 24 hours before debridement.
Paired-samples t-test (for parametric data) was used to compare the differences of inflammatory markers before and after surgery using the SPSS software (v22.0, IBM, USA).
Statistical significance was assumed if the p value was less than 0.05.

Results
A total of 27 patients were tracked from our registry. One patient died from gastric cancer at 18 months after surgery, and 2 patients lost to follow-up. These patients were excluded from the study leaving 24 included in the final analysis. 13  The combined bacteriology results were described in Figure 1, with an overall pathogen identification rate of 91.7% (22/24). The success rate for Staphylococcus, Streptococcus, and other pathogen was 100% (9/9), 71.4% (5/7), and 100% (6/6) respectively. Both the 2 failed cases were infected with Streptococcus agalactiae.

Discussion
The management of acute PJI presents great challenge to orthopaedic surgeons. To extract the implant and perform a one-stage or two-stage revision may be a hard choice for both the patients and surgeons. Thus DAIR, which retains the protheses, remains the mainstay treatment for acute PJI. However, its efficacy has been debated during the past few decades. Several studies have reported an overall success rate of lower than 50% [11,12,[14][15][16][17][21][22][23]. (Table 2) A systematic review by Silva NGS, especially mNGS, is an evolving technology widely applied in clinical diagnosis [35].
mNGS sequences all nucleic acid fragments in a clinical sample, enabling the use of bioinformatics method to obtain microbial sequences and species information and therefore identifies the pathogen. Recently, owing to the substantial cost reduction, mNGS has also been increasingly applied to the diagnosis of bone and joint infection [36]. Our previous reports have shown that mNGS has potential in identifying pathogens even in cases with antibiotic treatment and in cases with rarely-seen pathogens [37,38]. In this limited cohort of patients, mNGS also identified additional microorganisms in 2 cases, which added another evidence that mNGS may serve as a potential fast tool to diagnose PJI. However, more studies are warranted to prove its true efficacy. In our current practice, mNGS has been used as a routine diagnostic tool for suspected bone and joint infections.
Our second finding was DAIR could be considered even in patients whose symptoms were sustained for over 1 month.In our series, five cases were presented with symptom onset between 4 weeks to 8 weeks. These patients were all managed with DAIR and we saw no infection relapse during a minimum one-year follow up. A similar study was reported by Ottesen et al, which showed a high success rate of 88% in patients within 42 days [25].
Nonetheless, we have no experience in managing cases whose symptoms have been for over 3 months. Ottesen reported 10 such patients with DAIR with success rate of 60%, which was encouraging [25]. However, it merits further evaluation to see if DAIR could be performed to treat delayed or chronic PJI.
The third finding of our study was that we showed DAIR had high success rate for Staphylococcal infections. All the 3 cases with staphylococcus aureus infections were successfully treated, and even for the methicillin-resistant staphylococcus, as was seen in our 9 MRSE infections, DAIR completely eradicated infection. Staphylococcal infection has been viewed as a risk factor for failed DAIR, as literature reported high failure rate ranging from 45% to 76% (Table 3). One possible explanation is that Staphylococcus is more easier to form biofilm on the surface of implants, as well as its drug-resistance to antibiotics [3]. Thus several researchers have proposed staphylococcal infection to be a surgical contradiction to DAIR, especially for methicillin-resistant staphylococcus [39,40] [41]. For streptococcal infection, our limited data demonstrated a general satisfactory success rate of 71.4%(5/7), which was still higher than the data from a multi-center study (35%, 11/31) [15].
Another finding of this study is on the antibiotics regime. In our series, IV antibiotics were generally used for 2 weeks followed by 4 weeks of oral antibiotics, regardless of hip or knee infection. The 2013 guideline of Infectious Diseases Society of America (IDSA) has recommended IV antibiotics for 2-6 weeks followed by oral antibiotics for 3 months for hip infections, and 6 months for knee infections for treating acute PJI [44]. Nonetheless, the optimal duration of antibiotics has been controversial. Various studies have shown a similar infection eradication rate between short-term antibiotics and long-term antibiotics regime [34]. And a recent multi-center randomized trails did show that oral antibiotic therapy was noninferior to intravenous antibiotic therapy when used during the first 6 weeks for complex orthopedic infection [45]. Long-term antibiotics regime brings several adverse effects to the patients, and our study suggest a short-term antibiotics (  Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.  Bacteriology of included cases with positive culture/mNGS results.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.