Peri-Prosthetic Joint Infection after Finger Joint Arthroplasty

Abstract

Background:

Endoprosthetic treatment of finger joints is an established procedure. However, data on peri-prosthetic finger joint infections (PJI) are rare. The goal of this retrospective study was to assess infection rates and treatment outcomes after PJI of the finger joints.

Methods:

Between 1984 and 2014, 1,195 finger joints (978 metacarpal and 217 proximal interphalangeal [PIP] joints) in 518 patients were treated. A retrospective record analysis was conducted. Besides demographic data, infection rates, types of revision surgery, and treatment outcomes were evaluated. Treatment strategies changed during the treatment period. Routinely assessed patient-reported outcome measures, namely the QuickDASH score, visual analogue scale (VAS), an everyday function score (activities of daily living; ADL), and an adapted Clayton score, were compared.

Results:

In 36 finger joints of 30 patients, a post-operative peri-prosthetic infection could be identified (3%). Infections occurred on average 3.8 years after arthroplasty. For meta-carpal phalangeal (MCP) joints, the best clinical outcomes could be observed after a two-stage revision procedure (QuickDASH score 86). Arthrodesis performed in a one-stage procedure achieved favorable outcomes in PIP joints (QuickDASH score 89). Re-infection mandating revision surgery occurred in seven joints in five patients (7/36; 19.4%). In 25 patients with 29 finger joints, there was no re-infection and no need for revision.

Conclusion:

Peri-prosthetic infection of finger joint prostheses occurs in approximately 3% of all cases, which is a higher rate than in the more common hip and knee procedures. For MCP joint revision surgery, a two-stage procedure seems to be the best treatment choice. For PIP joint revision, a single-stage revision with arthrodesis in a functional position achieves a good outcome.

Long-term rheumatoid arthritis or osteoarthritis of the finger joints usually results in destruction of the joint articular surfaces. Both diseases also affect collateral ligaments, tendons, and peri-articular tissues. Finger deformations that often are painful as well as causing a limited range of motion (ROM) are the consequence. Joint ankyloses are likely. Isolated impaired joint function affects other fingers and finally limits the use of the entire hand.

Nowadays, endoprosthetic replacement of the metacarpal phalangeal (MCP) and proximal interphalangeal (PIP) joints is an established treatment to maintain range of motion (ROM) and good grip strength of the hand [1,2]. Different finger prosthesis designs are available for application. At present, the Swanson^® prosthesis (Wright Medical™, Memphis, TN, USA) is the gold standard for this purpose [2,3] Many studies have focused on long-term results for the Swanson^® prosthesis. After 10 years, radiologically obvious osteolysis is evident in 90% [4]. Implant fracture rates have been reported to be 30% after 10 years regardless of the location of the endoprosthetic joint replacement [5 –7]. Unfortunately, until now, the rates of peri-prosthetic finger joint infections (PJ) after arthroplasty have been published for only small groups or as case reports.

For hip and knee replacement surgery, PJI rates are less than 2% within 10 years after surgery [8,9]. Guidelines have been developed to improve treatment outcome. A two-stage procedure still is the gold standard for surgical therapy of hip and knee PJI [9,10]. Nonetheless, recurrence rates are high. Recently, for total knee arthroplasty, the recurrence rate of infection has been reported to be 19% after one year and 29% after six years [11]. Furthermore, it has been demonstrated that recurrence rates of infection depend on host characteristics as well as infection-causing pathogens [12,13]. For PJI after finger joint replacement, infection rates, rates of recurrence of infection, or functional outcome depending on treatment strategy have to be clarified. The aim of this study was to analyze the rates of PJI, re-infection rates after revision, and patient-reported outcomes depending on different treatment strategies.

Patients and Methods

The study was approved by the regional ethical review board of Rhineland-Palatinate, Mainz, Germany. All patients were fully informed about the study. They all gave written informed consent prior to study inclusion.

This retrospective analysis included all patients who underwent endoprosthetic replacement of the MCP and PIP joints between 1984 and 2014. All patients suffered from either rheumatoid/psoriatic arthritis (RA/PsA) or osteoarthritis (OA) of the finger joints. Risk factors for the peri-prosthetic infection are previous intra-articular injections of cortisol, skin defects close to the joint, and a difficult medical baseline therapy for RA with multiple medications. Patient age, underlying disease per se, or medication therapy after healing showed no effect on the incidence of infection.

A patient record analysis was performed. Patients who underwent finger joint replacement surgery were followed up in a standardized manner six weeks and annually after surgery.

In a case of PJI, intra-operatively harvested swabs or tissue samples were used for microbiologic analysis. Samples were cultured on blood agar plates. Incubation time at 37°C was seven days for aerobic and 14 days for anaerobic cultures. After pathogen identification, antimicrobial susceptibility testing was performed using standard techniques.

After collection of the tissue samples, antibiotic therapy was started empirically. Cefazolin has been used as the standard broad-spectrum antibiotic. In cases of a two-stage procedure, a gentamicin-loaded Septopal^® chain (Biomet, Freiburg, Germany) was inserted as a spacer. As soon as pathogens could be identified, antibiotic treatment was adapted to the antibiogram. In case of implant removal and debridement, antibiotic treatment was administered until clinical and laboratory signs of infection had disappeared. If arthrodesis and revision arthroplasty were both performed as a one-stage procedure, antibiotic coverage was carried out for six weeks. In two-stage revision procedures, antibiotic treatment was administered during the implant-free interval with an additional six weeks of antibiotic therapy after implantation of the new prosthesis. Beginning in 2006, rifampicin use was implemented to prevent biofilm formation after endoprosthetic joint replacement [12,14].

For follow-up, patients were questioned about pain using the visual analogue score (VAS) and their activities of daily living (ADL) limitations. Therefore, patient-reported outcome measures (PROM) were used pre- and post-operatively such as the disability measured by the arm, shoulder, and hand (QuickDASH) score, an ADL score according to the Hannover questionnaire of everyday function for rheumatic patients (Hannover Functional Questionnaire Backache [FFbH-R]; FFbH), and an adapted Clayton score (mobility, stability, ligament tension, and pain) [15]. The QuickDASH score, as an evaluation of the function of the upper extremity, was not available until 2002. Because most of the endoprosthetic finger joint replacement operations were performed before QuickDASH implementation, this score could be employed only partially. The QuickDASH and ADL scores do not distinguish between joints.

Data were analyzed using SPSS V. 25.0 (IBM, Armonk, NY). Collected datasets from clinical findings were explored for normality with descriptive statistics. The nonparametric Mann-Whitney U and Kolmogorov-Smirnov tests were used. Analysis of variance (ANOVA) with Bonferroni correction served as a control assessment as did the Games-Howell post hoc test. The critical value for significance was set at p < 0.05. Data were exhibited in graphs as means ± standard error of the mean (SEM).

Results

Five hundred eighteen patients could be included in the study. In total, 1,195 finger joint arthroplasties (978 MCP and 217 PIP) have been performed in those patients. The average age of the patients at the time of primary surgery was 59.3 (range 36.5–71.5) years. The period between the onset of complaints and revision surgery was a mean of 2.4 (range 0.17–5.0) years. The authors implanted Swanson endoprostheses almost exclusively; in only five cases were different prosthesis models implanted. The site of endoprosthetic joint replacement depended on the underlying disease. Rheumatoid arthritis (RA) was the predominant diagnosis for metacarpal involvement. In about 98% of cases, MCP replacement was related to RA. A PIP joint arthroplasty was performed in more than 50% of joints affected by osteoarthritis. The index finger was the site of the most often treated MCP, whereas endoprosthetic joint replacement of PIPs was most often performed on the middle finger.

Revision arthroplasty was necessary in 121 of 518 patients (23.4%). The main reason was recurrent malposition of the Swanson prosthesis with grip pathologies such as renewed ulna deviation and palmar subluxation and loss of ROM and grip power. The main reasons for mechanical malfunction were prosthesis failure and peri-prosthetic osteolysis. This osteolysis is caused by the pistoning effect of the silicone prosthesis with silicone synovitis on the bone interface. No skin perforation could be detected, as was previously described by Chopra et al. as a rare complication [16].

Peri-prosthetic infection occurred in 36 finger joints (3.0%; 30 MCP, 6 PIP). Revision surgery for infection was necessary after 3.8 (range 0.25–9) years after initial joint replacement. Of those patients, 21 with 26 affected joints could be recruited for follow-up. Six patients (seven joints) had already died. Three patients (three joints) were no longer identifiable or refused further contact with the department.

Two-thirds of the infections were caused by gram-positive cocci, in particular, Staphylococcus aureus. Other bacteria detected were Streptococci spp., Escherichia coli, Proteus mirabilis, and Pseudomonas aeruginosa.

An evolution of treatment could be seen when treatment strategies were evaluated. Between 1984 and 1996, a first group of 12 MCP joints were treated only with removal of the prosthesis and debridement. Seven patients could be followed up (mean 2.2 [range 0.5–5.9] years). The clinical results were poor in grip function, grip strength, and ROM. Earlier surveyed outcome values were adapted retrospectively to the QuickDASH score. It was found to be 112 points, which reflects a poor result. The Clayton score of 39 points (maximum possible 100 points) and in the ADL score with 19 points (maximum possible 40 points) could be found for this group (Table 1).

Table 1.

Infected Finger Prosthesis on MCP Joints by Treatment Regimen

	Group 1 (1984–1996)	Group 2 (1996–1999)	Group 3 (2000–2014)	Total
No.	7/12	6/8	8/10	30
Clayton Score	39	61.2	80.5
ADL Score	18.5	26.9	29.5
DASH Score	112	95	86.5
Follow-up (y)	2.2 (0.5–5.9)	5.9 (0.6–6.9)	2.4 (0.3–3.5)	2.5 (0.1–5)

Group 1 = sine-sine explantation; Group 2 = one-stage exchange, Group 3 = two-stage exchange.

In the second group of PJI after MCP joint replacement, a one-stage procedure with removal of the prosthesis, debridement, and immediate reimplantation was performed. In total, eight joints were treated in this manner between 1996 and 1999. Six could be followed up after 5.9 (range 0.6–6.9) years. A Clayton score of 61 points, an ADL score of 27 points, and a QuickDASH score of 95 points were determined during follow-up. Compared with the first group, in whom only removal and debridement was performed, all outcome scores were significantly better. However, poor results were found for patient-centered outcomes such as using a knife and fork, opening a bottle, or propping the head up on the hand (ADL 2.25/2.31 [range 2.25–2.83] points).

Since the Millennium, a two-staged treatment for PJI of MCP joints has been performed. In total, 10 MCP joint prosthesis were removed, and a gentamicin-loaded polymethylmethacrylate (PMMA) chain (Palacos^® Heraeus, Germany) was implanted. In addition to this application of local antibiotics, systemic antibiotic therapy was administered during the implant-free interval and additionally for six weeks (Fig. 1). Aspiration of the joint prior to reimplantation was carried out. Probably because of the small joints, no fluids could be obtained, and no microbiologic diagnostics could be performed prior to revision surgery. At 2.4 (range 0.3–3.5) years after reimplantation of an endoprosthesis, a Clayton score of 81 points, an ADL score of 30 points, and a QuickDASH score of almost 86 points could be determined (Table 1).

FIG. 1.

Therapy algorithm for finger joint infection. *Additional physiotherapy, ergotherapy, dynamic orthosis for tendon mobility; sonication (for detecting bacteria in biofilms on the prosthesis surface by ultrasonic preparation.)

In PJI of PIP joint prosthesis, removal and debridement were performed only between 1984 and 1986. The functional outcome was unsatisfactory. Hence, surgeons changed the treatment. A one-stage revision was performed with removal of the prosthesis, debridement, and consecutive arthrodesis in a functional position using k-wires. During follow-up of 2.5 (range 0.1–5) years, a Clayton score of 54 points, an ADL score of 26 points, and a QuickDASH score of 89 points was recorded in this group.

The results of the Mann-Whitney U test of all scores of the first group treated for MCP joint PJI were significantly worse than those of both other MCP groups, who underwent either one-stage or two-stage revision arthroplasty (p < 0.01). Comparing PIP arthrodesis for PJI treatment with the first group in which removal of the prosthesis and debridement was performed, better results could be documented for the arthrodesis group, as judged by ADL (p ≤ 0.018) and the other scores (p ≤ 0.001) (Fig. 2). All outcome scores of the second group (one-stage procedure) were significantly worse than those of the third group (p ≤ 0.015–0.001) but not significant compared with the PIP group for Clayton- and ADL scores (p ≤ 0.075/0.789). For the QuickDASH score, significance was seen (p ≤ 0.044). The third group (two-stage procedure) showed significantly better results than the PIP group in ADL and Clayton score (p ≤ 0.049/0.01) but not in in the QuickDASH score (p = 0.958) (see Table 2). However, the groups were small (MCP n = 30; PIP n = 6), so the sample was of limited significance.

FIG. 2.

Statistical analysis of ADL (A), Clayton (B), and DASH (C) score of post-operative results. MCP groups: Sine-sine situation (flail joint), one-time revision, and two-time revision. PIP group: Arthrodesis. Lines show statistically significant differences (p < 0.05).

Table 2.

Mann-Whitney U Test and Kolmogorov-Smirnov Test Comparing Results by Group

	Mann-Whitney U			Kolmogorov
	ADL	QuickDASH	Clayton	ADL	QuickDASH	Clayton
Group 1 vs. 2	0.000	0.000	0.000	0.000	0.001	0.001
Group 1 vs. 3	0.000	0.000	0.000	0.000	0.000	0.000
Group1 vs. PIP	0.018	0.001	0.001	0.008	0.002	0.001
Group 3 vs. 2	0.015	0.002	0.001	0.082	0.035	0.002
Group 2 vs. PIP	0.789	0.044	0.075	0.998	0.064	0.358
Group 3 vs. PIP	0.049	0.958	0.000	0.336	0.952	0.001

Nonsignificant results are marked in small italic type.

Group 1 = explant situation; Group 2 = one-stage exchange, Group 3 = two-stage exchange.

PIP = proximal inter-phalangeal.

The authors detected recurrence of PJI in four MCP joints in two patients of the first or second group. No case of reoccurring PJI was detected in the two-stage revision group. Both patients with recurrence had an early renewed infection of their one-stage revision of a directly implanted new Swanson^® prosthesis within six weeks.

Three revision operations were necessary in PIP joints after k-wire arthrodesis because of recurrence of infection (one to four months). One patient refused a stiffening or a flail joint. Thus, a two-stage exchange was performed. In one finger, re-arthrodesis was performed in a one-stage procedure. In one finger, progressive osteomyelitis occurred, necessitating finger amputation. Grip function of the single finger stump and the hand was worse but compensated for by the patient. In total, the rate of recurrent infection was 19% (7/36).

Discussion

The goal of this retrospective study was to assess the PJI and recurrence rates of PJI in finger joint prostheses. In addition, patient-reported treatment outcomes should be elucidated depending on different treatment strategies for finger joint replacement. The rates of PJI were higher than the rates after total hip and knee arthroplasty (3% after 3.8 years for finger joint prostheses versus <1.09% and 1.38% after five years for hip and knee arthroplasty, respectively) [8]. Achermann and coworkers demonstrated a PJI rate of 7.5% at an average of 45 months after elbow arthroplasty [1]. An association between basic medical treatment for inflammatory diseases and late infection could not be found. Poor soft-tissue coverage of the finger joints may be a reason for the higher infection rates compared with knee and hip PJI.

The infection-causing microorganisms play an essential role in PJI development. Evaluation of the microbial pattern of PJI in finger joint prosthesis revealed results similar to those of hip and knee PJI. In the present study, S. aureus was the most common PJI-causing pathogen, followed by Streptococcus spp. and less common (less than one third) gram-negative bacteria. Recently reported data from both osteosynthesis- and arthroplasty-related infections in hand surgery demonstrated S. aureus to be the most common microorganism (48% of 25 patients) [17]. For PJI after hip and knee arthroplasty, Staphylococcus spp. such as S. aureus and S. epidermidis were the pathogens mainly responsible, followed by Streptococcus and Enterococcus [18]. Rosteius et al. showed an increase in multi-resistant pathogens, but the data from our study did not demonstrate such a trend. This might be attributable to the long study period and the relatively low volume of finger PJI.

Recurrence of PJI was evident in 19.4% (7 of 36) cases. For recurrence of PJI in hip arthroplasties, Lange et al. evaluated 36 studies with a total of 1,304 patients in their meta-analysis. The average re-infection rate was 13.1% (95% confidence interval [CI] 10.0%–17.1%) for one-stage revision and 10.4% (95% CI 8.5%–12.7%) in the two-stage cohort [19]. After revision of elbow PJI, re-infection rates were as high as 30% [20]. Thus, also for recurrence, data for finger PJI were between the rates of hip and elbow PJI re-infections [21].

Different therapy regimens for PJI in finger joints and the clinical outcomes were assessed in the present study. The evolution of therapy was based on our experience with treatment strategies in hip and knee PJI treatment. To the best of our knowledge, the retrospective analysis of patient-reported outcomes depending on different treatment strategies for finger PJI is the first study taking PROMs into account. In the study of Meier et al., which analyzed implant-related infections of the hand, objective features such as infection-causing pathogens, treatment strategy, infection eradication, and duration of antibiotic therapy were assessed. Functional outcome was not determined, probably because of the inhomogeneous patient group [17]. Because no specific functional scoring system for fingers and finger joint replacement exists, it seemed reasonable to choose established evaluation methods such as the QuickDASH. Although infected finger joints might impair the function of the entire extremity, a bias could arise from patients being rheumatics. They often suffer from limitations of other upper-extremity joints, which has to be considered when interpreting the present results.

For MCP joints treated only by removal of the prosthesis with additional debridement, patient-reported outcomes were significantly worse than in both other groups (Fig. 2). In PIP joint prosthesis infections, better patient-reported outcomes were shown for implant removal, debridement, and one-stage PIP arthrodesis. According to PJI in hip and knee arthroplasty, the present data support considering a two-stage procedure in PJI of MCP joint infections to be the gold standard. For PJI in PIP joints, only removal and debridement and one-stage arthrodesis could be compared. Nevertheless, the present data are not able to answer the question of whether one-stage or two-stage re-arthroplasty of PIP joints leads to a better result.

The shortcomings of the present study are obvious. Being the first study focusing solely on PJI in finger joints, the volume of cases is low. To achieve a higher volume, multi-center studies would be necessary. Nevertheless, this retrospective study serves as a first step to focus on finger PJI and optimize patient-reported outcomes. National registry data, as performed for joints being more often replaced such as hip and knee, would be desirable for finger joint prosthesis as well [22].

Conclusion

After finger joint replacement, an infection rate of 3% was evident after 3.8 years (range 0.3–9 years) of follow-up. Recurrence of PJI appeared in 19%. Assessment of patient-reported outcomes after revision surgery of PJI in MCP joints revealed a two-stage approach to achieve favorable treatment outcomes. For PIP joints, primary arthrodesis is a valuable treatment option.

Footnotes

Acknowledgments

We thank Professor Heiner Thabe for the opportunity to give us his experience in the medical care of rheumatoid patients in general and in this study. He is and remains a great teacher for us.

We also are immensely grateful to Doctor Jochen Jung (Diakonie Krankenhaus Bad Kreuznach, kreuznacher diakonie) for his comments on an earlier version of the manuscript and lecture. Any remaining errors are our own and should not tarnish the reputations of this esteemed person.

CB designed the retrospective cohort study and was responsible for data collection. CB, MR, and GS were responsible for the patient-reported outcomes component of the study. CB and TE performed the statistical analysis and drafted the manuscript. CH, MR, VA, and GS corrected the manuscript to improve the quality of the written English. GS and TE made substantial contributions to analysis and interpretation of data. All authors read and approved the final manuscript.

The level of the evidence in this study is III.

Author Disclosure Statement

No competing financial interests exist.

References

Achermann

, Vogt

, Spormann

, et al. Characteristics and outcome of 27 elbow periprosthetic joint infections: Results from a 14-year cohort study of 358 elbow prostheses. Clin Microbiol Infect, 2011; 17:432–438.

Amornvit

, Rokaya

, Keawcharoen

, Thongpulsawasdi

. Stress distribution in implant retained finger prosthesis: A finite element study. J Clin Diagn Res, 2013; 7:2851–2854.

Bales

, Wall

, Stern

. Long-term results of Swanson silicone arthroplasty for proximal interphalangeal joint osteoarthritis. J Hand Surg, 2014; 39:455–461.

Chopra

, Jain

, Murali

, Chelliah

. A rare complication of a metacarpophalangeal joint replacement in a rheumatoid hand: A case report. Cases J, 2009; 2:7864.

Chung

, Kotsis

, Kim

. A prospective outcomes study of Swanson metacarpophalangeal joint arthroplasty for the rheumatoid hand. J Hand Surg, 2004; 29:646–653.

Clayton

. Surgical treatment at the wrist in rheumatoid arthritis: A review of thirty-seven patients. J Bone Joint Surg Am, 1965; 47:741–750.

Cochran

, Ong

, Lau

, et al.. Risk of reinfection after treatment of infected total knee arthroplasty. J Arthroplast, 2016; 31:156–161.

Goldfarb

, Dovan

. Rheumatoid arthritis: Silicone metacarpophalangeal joint arthroplasty indications, technique, and outcomes. Hand Clin, 2006; 22:177–182.

Kurtz

, Ong

, Lau

, et al. Prosthetic joint infection risk after TKA in the Medicare population. Clin Orthop Rel Res, 2010; 468:52–56.

10.

Lange

, Troelsen

, Thomsen

, Soballe

. Chronic infections in hip arthroplasties: Comparing risk of reinfection following one-stage and two-stage revision: A systematic review and meta-analysis. Clin Epidemiol, 2012; 4:57–73.

11.

Malchau

, Garellick

, Berry

, et al. Arthroplasty implant registries over the past five decades: Development, current, and future impact. J Orthop Res, 2018; 36:2319–2330.

12.

Meier

, Wirth

, Vogelin

, Sendi

. Characteristics and outcome of twenty-nine implant-related infections of the hand and fingers: Results from a twelve-year observational study. Surg Infect, 2018; 19:729–734

13.

Renz

, Perka

, Trampuz

. [Management of periprosthetic infections of the knee] (Ger). Der Orthop, 2016; 45:65–71.

14.

Rosteius

, Jansen

, Fehmer

, et al. Evaluating the microbial pattern of periprosthetic joint infections of the hip and knee. J Med Microbiol, 2018; 67:1608–1613.

15.

Schmidt

, Miehlke

, Witt

. [Status of the endoprosthesis in rheumatic metacarpophalangeal joints: Long-term results of metacarpophalangeal prostheses using Swanson's Silastic spacers] (Ger). Handchir Mikrochir Plast Chir, 1996; 28:254–264.

16.

Swanson

, de Groot Swanson

, Maupin

. Flexible implant arthroplasty of the radiocarpal joint: Surgical technique and long-term study. Clin Orthop Rel Res, 1984:94–106.

17.

Trampuz

, Piper

, Hanssen

, et al. Sonication of explanted prosthetic components in bags for diagnosis of prosthetic joint infection is associated with risk of contamination. J Clin Microbiol, 2006; 44:628–631.

18.

Trampuz

, Zimmerli

. Diagnosis and treatment of implant-associated septic arthritis and osteomyelitis. Curr Infect Dis Rep, 2008; 10:394–403.

19.

Triantafyllopoulos

, Memtsoudis

, Zhang

, et al. Periprosthetic infection recurrence after 2-stage exchange arthroplasty: Failure or fate?. J Arthroplast, 2017; 32:526–531.

20.

Wagner

, Luo

, Houdek

, et al. Revision proximal interphalangeal arthroplasty: An outcome analysis of 75 consecutive cases. J Hand Surg, 2015; 40:1949–1955.e1941.

21.

Yamamoto

, Malay

, Fujihara

, et al. A systematic review of different implants and approaches for proximal interphalangeal joint arthroplasty. Plast Reconstr Surg, 2017; 139:1139e–1151e.

22.

Zimmerli

, Trampuz

, Ochsner

. Prosthetic-joint infections. N Engl J Med, 2004; 351:1645–1654.