Introduction
The incidence of periprosthetic joint infection (PJI) of after THA is relatively low but increasing [8, 14 5
Recent studies identified some factors that were associated with re-revision, including reinfection after two-stage hip and knee exchange because of PJI. Many of the determined factors were related to patient morbidity, but some pathogens were considered to be associated with the inability to eradicate the infection [3, 13 15
One-stage revision arthroplasty to manage PJI provides several benefits with respect to patient morbidity and mortality as well as socioeconomic aspects. Because one stage-exchange has superior functional outcomes compared with two-stage exchange, with similar reinfection rates, this approach is becoming more popular among orthopaedic surgeons [9
To date, there are some recommended contraindications of a one-stage revision arthroplasty, such as the preoperative failure to isolate the microorganism or more than two unsuccessful one-stage procedures [4, 19 4, 10, 19
Therefore, we asked: after one-stage revision arthroplasty for PJI of the hip, what factors were associated with an increased risk of revision, and what factors were associated with an increased risk of reinfection?
Patients and Methods
Study Design and Patients
After obtaining approval to perform the study from our institutional review board, we performed this retrospective case-control study at our arthroplasty center. In our available electronic database from 2009 to 2017, we searched for patients who underwent revision for any reason after one-stage revision arthroplasty for PJI of the hip at our hospital.
In the first 3 years of the study, PJI was diagnosed when a pathogen had been isolated from at least two separate tissue or fluid samples or in cases when the sinus tract was communicating with the joint. Later on, the criteria of the Musculoskeletal Infection Society were considered [12 11
Generally, the one-stage procedure is performed at our institution as long as the causative microorganism is known and the patient does not have a concurrent sepsis. However, based on the intraoperative findings, such as the extension of bone and soft tissue infection or the involvement of neurovascular structures, the decision for a planned one-stage exchange can be changed. For inclusion in the “case” group in this case-control study, patients were required to have complete documentation of the one-stage revision arthroplasty for PJI of the hip and one or more subsequent revisions. The control group (121 patients) consisted of patients who were 1:1 matched to patients in the re-revision group by year of birth, sex and year of the one-stage hip revision. Similar to the re-revision patients, controls were treated for PJI of the hip with one-stage revision arthroplasty; however, they had not undergone subsequent revision for any reason by the latest followup examination. The mean followup of the control group was 66 months (range, 17-119 months).
We performed all procedures following the standard protocol of our hospital using prepared antibiotic-loaded bone cement. After radically débriding infected and necrotic tissue, including bone, and completely removing the prosthesis and cement, chlorhexidine-based soaks were placed while a sterile re-draping was performed before reimplantation of the cemented components [4
The endpoint of interest was any revision procedure after one-stage direct exchange revision for PJI of the hip, defined as any reoperation requiring revision of one or more components of the hip arthroplasty including irrigation and débridement. Reinfection was defined as any subsequent PJI with the same or new pathogen.
Demographics
There were 66 men (55%) and 55 women (45%) from a total of 121 patients in each group. The mean age of the patients with any re-revision surgery was 67.5 years (SD, 9.7 years) with a mean BMI of 29.1 kg/m2 (SD, 5.7), and that of patients included in the control group without any repeat revision surgery at the latest followup examination was 68.2 years (SD, 9.0 years) with a mean BMI of 28.1 kg/m2 (SD, 5.5 kg/m2 ). Preoperatively, patients with re-revisions and the controls had a mean C-reactive protein (CRP) value of 40.7 mg/L (SD, 50.0 mg/L) and 30.2 mg/L (SD, 34.7), respectively. The mean length of hospital stay was 26.2 days in the re-revised group (SD, 11.6 days) who had a mean of 4.6 (SD, 3.1) prior surgeries of the investigated hip, compared with 22 days for the controls (SD, 6.6 days) who had a mean of 2.8 (SD, 2.1) prior surgeries (Table 1 ).
Table 1.: Demographic data and baseline characteristics of patients in both groups
Timing of and Indications for Re-revision
The re-revision procedures were performed after a mean of 407 days (range, 14 to 2599 days). The most common causes of any re-revision surgery were repeated hip dislocation in 53 of 121 patients (44%), reinfection in 40 of 121 patients (33%), and aseptic loosening in 16 of 121 patients (13%) (Table 2 ). Cemented polyethylene with a head size of 32 mm and dual mobility cups with a head size of 28 mm were most commonly used (Table 3 ). Most reinfections were new infections caused by newly identified microorganisms (n = 23 of 40; 58%), whereas recurrent reinfection with the same bacterium occurred in 15 of 40 patients (38%). In two of 40 patients, the pathogen could not be identified. Among all patients in the re-revised group (n = 121), 43 (36%) underwent another re-revision procedure after the initial one-stage index revision arthroplasty of the hip for the following reasons: reinfection (n = 19), dislocation (n = 15), wound revision (n = 4), aseptic loosening (n = 3), and periprosthetic fracture (n = 2).
Table 2.: Indications for re-revision
Table 3.: Details and numbers of the implanted prosthesis components in all patients
Investigated Factors
We performed univariate analyses including more than 40 patient-, joint-, and infection-related factors for both repeat revision and reinfection after the one-stage revision hip arthroplasty for PJI. Factors with p values < 0.1 were included in the multivariate analysis.
Patient-related factors included age, sex, BMI (kg/m2 ), weight, smoking status, and preexisting comorbidities, such as diabetes mellitus, coronary heart disease, hyperlipidemia, chronic obstructive pulmonary disease, liver disease, renal failure, thyroid disease, rheumatoid arthritis, depression, dementia, tumor history, and history of deep vein thrombosis or pulmonary embolism. Regarding the affected joint and surgery, the following factors were determined: the number of prior surgeries, previous procedure due to infection (débridement and irrigation, one-stage, or two-stage exchange), bacteria identified from any prior surgery due to infection, pattern of previous PJI (monomicrobial or polymicrobial), bacteria identified during the one-stage procedure, the presence of a sinus tract, duration of surgery, wound closure (staples versus suture), postoperative wound healing disorders, and persistent wound drainage, defined as a prolonged secretion for longer than one week. Preoperative serum laboratory parameters such as white blood cell count, CRP, potassium, sodium, glucose, and hemoglobin levels were included in the analysis. Postoperative urinary tract infection, pneumonia, acute renal failure, deep vein thrombosis, pulmonary embolism during the hospital stay, allogeneic blood transfusion, and the length of hospital stay were also investigated. Factors with p values < 0.1 were further advanced to be included in the multivariate analysis.
Statistical Analysis
We processed data and performed the statistical analyses using SAS 9.3 (SAS Institute Inc, Cary, NC, USA). Descriptive statistics are demonstrated for both groups for the number of occurrences and percentage, mean, SD, and extrema. The Shapiro-Wilk test was used to examine if the data were normally distributed. If the data demonstrated a normal Gaussian distribution, we performed a t-test. Otherwise, we used the Mann-Whitney test. The frequency distribution of categorical variables was compared between the two groups using Fisher's exact test. All tests were two-tailed. For selected variables, we calculated odds ratios (ORs). Using a stepwise method, we also performed a multivariate logistic regression analysis of variables with variables with p values less than 0.1, including length of hospital stay, preoperative CRP, prior surgery due to infection, history of polymicrobial PJI, prior one-stage or two-stage revision arthroplasty, coronary heart disease, sinus tract infection, prolonged wound drainage, previous PJI with Staphylococcus epidermidis and isolation of enterococci during the index surgery. All analyses were repeated to include only patients who underwent re-revision for reinfection, and these patients were then compared with those in the control group. In addition to the included variables in the multivariate analysis of re-revision for any reason, the following factors were included in the multivariate analysis for reinfection: BMI, liver disease, and renal disease.
Results
The independent factors associated with repeat revision for any reason after one-stage revision arthroplasty were prolonged wound drainage beyond 1 week (OR, 7.4; 95% CI, 2.6–20.6; p < 0.001), intraoperative isolation of enterococci (OR, 4.8; 95% CI, 1.4–15.7; p = 0.010), and prior surgical procedure due to infection of the index hip (OR, 3.6; 95% CI, 2.0–6.4; p < 0.001) (Table 4 ).
Table 4.: Independent risk factors of failure for any reason and reinfection according to the multivariate logistic regression analysis
The factors associated with reinfection including both new and persistent infections were prolonged wound drainage (OR, 6.9; 95% CI, 2.2–21.5; p = 0.001) and prior surgery due to infection (OR, 4.3; 95% CI, 1.9–9.5; p < 0.001) (Table 4 ).
Discussion
PJI is managed mainly using the one- or two-stage exchange. The general condition of the patient, the extent of bone and soft tissue damage and the causative microorganisms influence the preoperative decision for the treatment approach [5 3, 13, 15 Enterococcus were independently associated with repeat revision for any reason after the one-stage revision for PJI. Prolonged wound drainage and prior surgery for infection of the affected hip remained also independently associated with re-revision due to repeat infection.
Our findings should be interpreted in light of some important limitations. First, the retrospective design might have been associated with selection bias. Although we strived to capture all patients who underwent re-revisions after one-stage revision arthroplasty of the hip for PJI, other patients might have undergone further revisions at external hospitals and should be considered lost to followup. Accordingly, the re-revision causes and the identified factors as well might have been influenced. Similarly, the probability of subsequent revisions among the controls at external hospitals is a further limitation. However, as a tertiary referral center we are regularly contacted regarding cases with postoperative complications after the one-stage exchange due to the complexity of the procedures. Second, although the mean time to repeat revision of the case group was shorter than the minimum duration of followup of the control group, the overall duration of followup is relatively short, and longer followup of patients who have served as a control group might reveal the need of re-revisions among them, which restrict the eligibility to be included as controls. Third, regarding our statistical analyses, the current matched design and not using pair-matched study design such as a conditional logistic regression analysis limits our results. Unfortunately, only about 36% of the case-control orthopaedic studies used appropriately matched methods [7
Prolonged wound drainage, infection with enterococci, and previous hip surgery for infection were independent factors associated with repeat revision after one-stage revision arthroplasty for PJI. An important factor in all analyses was prolonged wound drainage, which was associated with an approximately sevenfold higher risk of re-revision and reinfection. Persistent wound drainage beyond 1 week postoperatively was considered prolonged wound drainage and may be a sign of developing infection after joint arthroplasty. Therefore, early surgical treatment is advocated when drainage persists beyond the fifth to seventh day postoperatively [18
Likewise, patients with prior surgery due to infection of the revised hip were at higher risk of any re-revision, particularly because of reinfection with an approximately 3.5- and fourfold, respectively. In the most extensive series to date about the results of one-stage revision arthroplasty for PJI of the hip, Buchholz et al. [2 1 13, 17
With respect to the isolated pathogen, Enterococcus species were associated with re-revision for any reason in our study; they are known antibiotic-resistant opportunistic pathogens and are classified as difficult to treat [20 6, 16
Being the most common cause of re-revision, repeated dislocation could be explained by the performed radical débridement, and if necessary with sacrificing of basic anatomical structure for hip stability. In our analysis of the prosthesis components used for re-implantation, the size of the prosthesis head and the use of modular stems did not affect the rate of re-revision because of repeated dislocation. However, the reimplantation of a dual mobility cup was associated with a lower postoperative hip dislocation.
Although we identified a relatively large number of patients who underwent any re-revisions after one-stage revision arthroplasty for PJI of the hip, reinfections accounted for only one-third of the total number of re-revisions. However, we primarily analyzed the host and local risk factors leading to re-revision for any reason, including reinfection, because the goal of one-stage revision arthroplasty should be not only to control infection but also to achieve an acceptable re-revision-free survival rate that reflects better functional outcome.
In conclusion, we identified three nonmodifiable factors that showed independent associations with re-revision after one-stage revision arthroplasty for PJI of the hip. Prolonged wound drainage after the one-stage revision arthroplasty for PJI of the hip must be treated rigorously. Patients with a history of prior surgical procedure due to infection of the hip should be informed about the risk of further re-revision when considering the one-stage exchange. In case of enterococcal infection, surgeons may consider another treatment approach rather than the one-stage exchange. However, further studies comparing one-stage and two-stage revision arthroplasty and considering the limitations in the current study are required to determine specific risk factors of repeat revision for each of both management strategy. Lastly, because instability is the most common cause of repeat revision, we recommend the use of dual-mobility cups when performing the one-stage revision arthroplasty of the hip.
Acknowledgments
We would like to thank Dr. Silke Lange for the performance of the statistical analyses in this study.
References
1. Brown TS, Fehring KA, Ollivier M, Mabry TM, Hanssen AD, Abdel MP. Repeat two-stage exchange arthroplasty for prosthetic hip re-infection. Bone Joint J. 2018;100:1157-1161.
2. Buchholz HW, Elson RA, Engelbrecht E, Lodenkämper H, Röttger J, Siegel A. Management of deep infection of total hip replacement. J Bone Joint Surg Br. 1981;63:342-353.
3. Fagotti L, Tatka J, Salles MJC, Queiroz MC. Risk factors and treatment options for failure of a two-stage exchange. Curr Rev Musculoskelet Med. 2018;11:420-427.
4. Gehrke T, Kendoff D. Peri-prosthetic hip infections: in favour of one-stage. Hip Int. 2012;22(Suppl 8):S40-S45.
5. Jiranek WA, Waligora AC, Hess SR, Golladay GL. Surgical Treatment of Prosthetic Joint Infections of the Hip and Knee: Changing Paradigms? J Arthroplasty. 2015;30:912-918.
6. Kheir MM, Tan TL, Higuera C, George J, Della Valle CJ, Shen M, Parvizi J. Periprosthetic joint infections caused by Enterococci have poor outcomes. J Arthroplasty
. 2017;32:933-947.
7. LeBrun DG, Tran T, Wypij D, Kocher MS. How often do orthopaedic matched case-control studies use matched methods? A review of methodological quality. Clin Orthop Relat Res. 2019;477:655-662.
8. Lenguerrand E, Whitehouse M R, Beswick A D, Jones S A, Porter M L, Blom A W. Revision for prosthetic joint infection following hip arthroplasty. Bone & Joint Research
, 2017;6:391-398.
9. Leonard HAC, Liddle AD, Burke O, Murray DW, Pandit H. Single- or two-stage revision for infected total hip arthroplasty? A systematic review of the literature. Clin Orthop Relat Res. 2014;472:1036–1042.
10. Nguyen M, Sukeik M, Zahar A, Nizam I, Haddad FS. One-stage exchange arthroplasty for periprosthetic hip and knee joint infections. Open Orthop J
. 2016;10:646-653.
11. Parvizi J, Gehrke T, Chen AF. Proceedings of the International Consensus on Periprosthetic Joint Infection. Bone Joint J. 2013;95-B:1450-1452.
12. Parvizi J, Zmistowski B, Berbari EF, Bauer TW, Springer BD, Della Valle CJ, Garvin KL, Wongworawat MD, Fehring TK, Osmon DR, Barrack RL, Berend KR, Esterhai JL, Gioe TJ, Kurtz SM, Masri BA, Spangehl MJ, Segreti J. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res. 2011;469:2992-2994.
13. Schwarzkopf R, Mikhael B, Wright E, Estok DM, Katz JN. Treatment failure among infected periprosthetic total hip arthroplasty patients. Open Orthop J. 2014;8:118–124.
14. Springer B D, Cahue S, Etkin C D, Lewallen D G, Mcgrory B J. Infection burden in total hip and knee arthroplasties: An international registry-based perspective. Arthroplasty Today. 2017;3:137-140.
15. Staats K, Boehler C, Frenzel S, Puchner SE, Holinka J, Windhager R. Failed two-stage exchange: factors leading to unachievable endoprosthetic reconstruction after multiple revision surgeries. J Arthroplasty. 2018;33:195-199.
16. Tornero E, Senneville E, Euba G, Petersdorf S, Rodriguez-Pardo D, Lakatos B, Ferrari MC, Pilares M, Bahamonde A, Trebse R, Benito N, Sorli L, del Toro MD, Baraiaetxaburu JM, Ramos A, Riera M, Jover-Sáenz A, Palomino J, Ariza J, Soriano A; European Society Group of Infections on Artificial Implants (ESGIAI). Characteristics of prosthetic joint infections due to Enterococcus sp. and predictors of failure: a multi-national study. Clin Microbiol Infect. 2014;20:1219-1224.
17. Triantafyllopoulos GK, Memtsoudis SG, Zhang W, Ma Y, Sculco TP, Poultsides LA. Periprosthetic infection recurrence after 2-stage exchange arthroplasty: failure or fate? J Arthroplasty. 2017;32:526-531.
18. Wagenaar FBM, Löwik CAM, Zahar A, Jutte PC, Gehrke T, Parvizi J. Persistent wound drainage after total joint arthroplasty: a narrative review. J Arthroplasty. 2019;23:175-182.
19. Zahar A, Webb J, Gehrke T, Kendoff D. One-stage exchange for prosthetic joint infection of the hip. Hip Int. 2015;25:301-307.
20. Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. N Engl J Med
. 2004 Oct 14;351:1645-54.
