Infection after shoulder replacement: A literature review and strategies on diagnosis, management, and clinical outcomes

Abstract

Infection following shoulder arthroplasty is on the rise as recent data have shown, potentially due to the increase in the number of shoulder arthroplasties being performed. Despite this increase, infective complications in lower limb procedure still exceed those of the upper limb. Infection in the shoulder usually occurs as a low-grade infection due to the most common causative pathogen (Propionibacterium acnes) and can be easily missed. Therefore, a high index of suspicion is needed to make the diagnosis. Furthermore, the use of MSIS criteria can greatly enhance the diagnosis of Periprosthetic Joint Infection (PJI) if appropriately utilised. There are a number of options in the treatment of PJIs; with the 2-stage procedure appearing to be the most commonly utilised, however, the 1-stage has been shown to have similar outcomes, if not superior. The choice of antibiotics should be determined by the antimicrobial susceptibility test and in conjunction with local guidelines and microbiology support. This is usually done as part of the preoperative work up from the pre-surgical aspirate/specimen cultures.

1. Introduction

As the number of shoulder arthroplasties performed increases every year, so does the complications following shoulder arthroplasty. They have become an increasing global concern, and of these various post-operative complications, infections contribute to the bulk of morbidity experienced with the procedure. Post-operative shoulder infection, although uncommon, have been reported to be in the range of 0.61–2.5 %.^1,2 Just as in lower limb PJIs, this has a substantial public health and financial impact resulting in reduced quality of life and increased burden on healthcare systems. Recent review by Faria et al. revealed that the mean hospital cost for a 2-stage revision surgery following prosthetic infection of the shoulder was approximately $35,824.³ The number of shoulder arthroplasties performed is on the rise as shown by data from the United states where there has been a rapid surge in the number of shoulder and elbow joint replacements from 28,000 per year between 2000 and 2008 to about 100,000 in the year 2015,² and this is expected to increase seven fold over the next 15years. This is expected to impact on the overall number of complications, including a corresponding rise in PJIs. Nezwek et al. noted in a review an increase in number of postoperative complications following reverse Total Shoulder Arthroplasty (rTSA) in the United State with an incidence of revision surgery from 3.3 % to 10.1 % each year and this is attributed to the increase in the number of rTSAs performed yearly.⁴ Primary rTSA appears to be associated with an increased risk of infection as compared to primary anatomical total shoulder arthroplasty (aTSA) across a number of small size studies. However, other confounders such as patient factors, demographics, co-morbidities and social history may make this conclusion less certain. For example, in a retrospective cohort study by Knapp et al. recruiting 12,530 candidates with 6949 (54.4 %) female and 5499 (44.5 %) male, it was found that more male patients have an increased superficial surgical site infection with Odd Ratio (OR) of 2.63, p = 0.035 but overall infection rate is similar in both groups.⁵ Given the proximity to the armpit, some have hypothesized that this may contribute to why the main causative organism is Propionibacterium acnes, an organism found in the dense hairy regions and upper part of the body, and that males are particularly at higher risk of infection by this organism. The aim of this review is to look into Infection following shoulder arthroplasty and to discuss new concepts and management protocols in tackling the issue of PJIs.

2. Methodology

An online literature search was conducted included databases such as Google Scholar, PubMed Central, Biomed Research International with a focus on articles published in the last five years 2018–2023. Studies employing innovative strategies for infection prevention and management were prioritised. The key search terms were: Upper limb surgeries, infection after shoulder surgeries, incidence of infection complicating shoulder surgeries, volume of upper limb surgeries vs lower limb surgeries, causative organisms and risk factors for infection after shoulder surgeries, Diagnostic techniques and treatment of infection following shoulder arthroplasty. Studies were excluded if they were in languages other than English (Fig. 1).

Fig. 1.

Flow diagram for search and selection of included paper reference.

3. Literature review

Shoulder arthroplasty PJIs are less common than lower limb PJIs. This likely results from the fact that fewer shoulder and elbow arthroplasties by absolute numbers are carried out compared to hip and knee arthroplasty.⁶ Although there seems to be a trend showing a gradual increase in the numbers of upper limb arthroplasty being carried out, it is still significantly lower in absolute numbers than its lower limb counterpart. By simple association, with increased upper limb arthroplasty we would expect to see an increase in post-operative complications in general, with an expected rise in PJIs. For instance, in a study by Valsamis et al. it was shown that the amount of shoulder replacement surgeries done in England between 1999 and 2020 in NHS hospitals and NHS-funded care quadrupled from 2.6 to 10.4 per 100,000 population and with it, an increased infection rate. This number may even be up to 10 % higher as the data does not capture privately funded activities undertaken in the independent hospitals, and data from the National Joint Registry.⁷ Also, Data from the Italian Register of Arthroplasty (RIAP) from 2018 to 2020 according to Romano et al. revealed an increase of 18.3 % in shoulder replacement surgeries from previous.⁸

Apart from the increased volume of shoulder arthroplasty performed in recent years, other factors have been proposed to have been contributory to the rise in infection. This include whether the surgical procedure is a prior prosthetic implant, male factor (2.5 times higher than females) and young age, increase duration of surgical procedure.⁹

4. Diagnosis

Effective management of any infection results from well-orchestrated efforts and assessment made from history taking, meticulous examination, appropriate investigations and formulating a diagnosis. Some of the physical signs noticeable with shoulder joint PJIs include: swelling, redness, sinus tract, pain and restriction of movement of the shoulder joint (Fig. 2). However, making a diagnosis of infection after shoulder arthroplasty may be very daunting as this is relatively uncommon compared to its more established lower limb counterpart. Therefore, a high index of suspicion is required, especially considering the sometime indolent nature. The main causative organisms are skin flora (Propionibacterium acnes and Staph sp.), and the clinical presentation may also be nonspecific. Propionibacterium acnes is a low virulent organism and so infection does not always elicit a significant systemic reaction. Major symptoms of presentation are pain and joint stiffness, and this can be from other non-infective complications. The 2018 International Consensus Meeting (ICM) on Orthopaedic Infections created a scoring system to diagnose shoulder prosthetic joint infections. They grouped the various factors into major and minor criteria. Based on the score it was classified as Definitive infection, Probable infection, Possible infection and Unlikely infection. Table 1, Table 2 shows the components of each criterion with score of ≥ 6 showing definitive infection.¹⁰

Fig. 2.

Photograph of left shoulder joint infection with evidence of inflammation and abscess formation.

Table 1.

International Consensus Meeting (ICM) scoring system for diagnosis of shoulder PJI.¹⁰.

	Description	Weight/Score
Major Criteria	a)Presence of sinus tract b)2 positive tissue cultures with phenotypically identical virulent organisms c)Gross intra-articulate pus	Definite PJI – meeting one of the 3 major criteria
Minor Criteria	Unexpected wound drainage	4
	Single positive tissue culture with virulent organisms	3
	Single positive tissue culture with low virulent organism	1
	Second positive tissue culture (identical low virulent organism)	3
	Numeral loosening	3
	Positive frozen section (5 PMNs in >/ = 5 high-power fields)	3
	Positive preoperative aspirate culture (low or high virulence)	3
	Elevated synovial neutrophil percentage (>80 %)	2
	Elevated synovial WBC count (>3000 cells/microlitre)	2
	Elevated ESR (>30 mm/h)	2
	Elevated CRP level (>10 mg/L)	2
	Elevated synovial alpha-defensin level	2
	Cloudy fluid	2

Table 2.

ICM definition of diagnosis of Periprosthetic Joint Infection (PJI)¹⁰.

Definite PJI	One of three major criteria present
Probable PJI	>/ = 6 minor criteria with identified organism
Possible PJI	a)≥ 6 minor criteria without identified organism OR b)6 minor criteria with positive cultures with virulence organism (e.g: MSSA, MRSA) OR c)6 minor criteria with 2 positive cultures with low virulence orgaganism (e.g: C.acnes, Fungus)
Unlikely PJI	<6with negative or only single positive culture with low virulence organism

Another investigation that may help with diagnosis includes the use of radiological techniques. The initial imaging technique used is the plain radiographs of the shoulder joint (Fig. 3). Signs that may suggest infection are areas of radiolucency at the bone-metal interface or cement-bone interface depicting osteolytic lesion. This may, however, be present in aseptic loosening of the prosthesis and therefore, it is an inconclusive test and has low sensitivity. Bone scintigraphy with 99mTc has excellent sensitivity but its specificity for PJI is very low as positive uptake of the radioactive material is normally present even up to 2 years after implantation without any infection. The use of MRI and CT scan is limited due to cost and low specificity. Though many contemporary images are being deployed in the diagnosis of PJI, the International Consensus Meeting (ICM) has not included imaging as part of diagnostic criteria for PJIs. The American Academy of Orthopaedic Surgeons (AAOS) has combined these tools into an algorithm to make diagnosis of an infected joint (Fig. 4).¹¹

Fig. 3.

Plain radiographs showing periprosthetic radiolucency and bone resorption with loosening of the prosthesis as in image B as depicted by prosthesis malalignment following Joint infection.

Fig. 4.

American Academy of Orthopaedic Surgeons's Algorithm for diagnosis of Prosthetic joint infections.¹¹.

Biological markers such as, WBC, C- reactive proteins, Erythrocyte Sedimentation Rate, Procalcitonin (PCT), have been employed in the diagnosis of PJIs and hence, they are evident in the ICM criteria for diagnosis. Though they are categorised as minor criteria, they help in improving the accuracy of diagnosis. Newer evidences are pointing to the use of Alpha-defensin, an endogenous peptide released by the polymorphonuclear leukocytes in response to pathogens invasion in the synovial fluid and consistent results for the diagnosis of PJIs have been observed using this biomarker regardless of which types of organism is the causal agent.¹²

Molecular studies using the Polymerase Chain Reaction (PCR) techniques is the emerging diagnostic approach that is been employed. This technology uses the knowledge of the amplification of the genetic component of organisms to grow them in the lab and this can be used for an otherwise unculturable PJIs. A specific quantitative PCR (qPCR) assay can be used to perform a diagnostic test identifying a particular bacteria species (e.g: S. aureus), group of closely related species (e.g: Staphylococcal species) or a common resistance genetic make up (mecA as in Methicillin Ressitant S. aureus (MRSA)). On the other hand, broad-range PCR such as 16S rRNA when use will require additional diagnostic method in order to identify the specific bacteria cause. These additional methods include cloning PCR amplicons and sequence analysis using the next-generation sequencing techniques.¹²

5. Treatment

Prevention of this complication following shoulder arthroplasty is both time and energy saving compared to treatment of the infection. Finance is also substantially preserved in prevention compared to treatment ensueing from the complications. As part of the preventive measures, observing universal sterile techniques in surgery by all staff members undertaking the procedure is key. Making sure the prostheses used are within the expiry period and that the sterile seals are intact also contribute to the prevention of infection. While the above may be observed keenly, post operative infections following shoulder arthroplasty may not be totally eliminated and in cases when this occurs some treatment options are available.

The standard treatment of infection after shoulder arthroplasty is revision surgery. Treatment with only antimicrobial agents reduces the chance of adequate infection resolution or worsening symptoms. Ultimately this increases the chance of antibiotics resistance of the organism due to prolonged unnecessary exposure to such antimicrobial agents.¹²

PJIs are caused by a plethora of bacterial species with different degree of virulence. Virulent bacterial PJIs such as from S. aureus tend to have early/acute onset with presentation like joint effusion, fever and leukocytosis while the less virulent ones such as C. acne are associated with more subtle and chronic symptoms. There is no consensus time limit for determining the chronicity or acuity of PJIs but in a study by Talsma et al., a period of 3months has been used with presentation within 3months of initial arthroplasty termed acute/early PJIs and presentation after that period termed chronic PJIs.¹³

In the treatment of early post operative shoulder PJI, radical surgical debridement of the periprosthetic tissue and synovectomy is an option. This is done with thorough irrigation of antiseptic fluid of the tissue in an open procedure with inlay prostheses exchange performed. Fink et al. noted that irrigation should not be done arthroscopically as this has a lower success rate compared to open procedures. With this approach to management, antibiotic sensitivity are not usually available and empirical antibiotics combinations are usually started before sensitivity results are known. Zimmerli et al. advocated for the use of rifampicin due to its activity against non-resistant bacterial biofilm. Zimmerli achieved a 100 % success rate after treatment of 12 periprosthetic infection with Rifampicin and Ciprofloxacin combination as compared to 58 % with ciprofloxacin treatment with placebo. Whereas Berdal reported 82 % success in treatment of 29 patients with combination of rifampicin and ciprofloxacin.¹² This is due to the ability of rifampicin to target sensitive Gram-positive pathogens in the biofilms and the fluoroquinolone (e.g: Ciprofloxacin) targeting the Gram-negative pathogens. Other combinations of antibiotics have been used, including Rifampicin with Vancomycin, in target populations that have fluoroquinolone resistance.

In chronic infection the options for treatment are a single-stage revision, two-stage revision, use of a permanent spacer and lifelong use of suppressive antibiotics. The choice will hinge on patient factors, surgical decision making and patient choice, considered on a case-by-case basis.

A single-stage revision (SSR) involves extensive surgical debridement of the periprosthetic tissue and radical synovectomy. This is routinely followed by a thorough irrigation of tissue before a new implant is introduced with antibiotic-impregnated cement. In order to effectively perform this procedure, it is imperative that the culture test results be available, and sensitivity confirmed so as to determine the correct and appropriate choice of antibiotics to use for the cement. Studies using this concept have achieved infection-free survival of between 90 % and 100 %.¹¹ The advantages of this procedure are time and cost saving and lack of need for patients to undergo another surgeries with its impending anaesthetic complications, however, the disadvantages include issues associated with spacers such as fracture, bone resorption and limb shortening. However, the success of this procedure is greatly enhanced if the causative infective agent and its antibiotic sensitivity is known prior to the procedure. In some studies, this procedure has proven to give comparable outcome and in others an improved outcomes as compared to the two staged procedures.¹⁴

A 2-stage revision (2-SR) operation is the most common method used for the treatment of shoulder PJIs. The first procedure entails removal of the infected prosthesis and thorough irrigation and debridement with insertion of an antibiotic spacer in the joint to prevent excessive fibrosis and stiffness whilst awaiting clearance of infection and the second stage. At the first stage, biopsies are taken for culture and histology. During the second stage, it is imperative that culture and sensitivities are available and the antibiotics specific to the pathogen isolated are prescribed in the impregnated cement mixture and with a new implant in situ. This method is suitable for when MSIS score is inconclusive and having no culture results available in the first instance. Usually with this procedure, the rotator cuff mechanism is often inadequate following debridement, and a reverse shoulder prosthesis is advisable if the initial index case had been an anatomic prosthesis. Alexander et al. concluded in their review that this procedure resulted in achievement of greater postoperative functionality as measured by the American Shoulder and Elbow surgeons (ASES) and the Simple Shoulder Test (SST) scoring systems, but the overall lower infection clearance of 85.2 % as compared to 95.6 % for a 1-stage procedure was noticed.¹⁴

Table 3 shows the results of various studies comparing the outcomes of single stage and 2-stage revision for shoulder joint infections. While older studies showed no significant differences between the two groups, recent studies have shown one stage revision to have better functional outcomes and lesser re-infection rate. However this has to be interpreted with caution keeping in mind the selection bias associated with the individual studies. Most of the studies included were retroscpective studies with no clear selection criteria for single and two stage revision. Hence, a biased selection of the treatment algorithm such as one-stage revision in simple cases and two-stage revision in complex cases could not be completely ruled out. Based on available data, we can conclude that under the right conditions, 1-stage revision provides better results in infection control, with lower complications and possible better clinical outcomes.

Table 3.

Studies comparing outcomes of the single (SSR) vs 2-stage Revision (2-SR) procedure for shoulder PJIs.

Author	Study type	Variables measured	Outcome		p-value	Comments
			SSR	2-SR
Gregory et al.15 2016	Systematic Review of 30 studies	Functional Outcome (Average Constant Murley score)	48.1	40.9	0.06	No difference between SSR and 2-SR
		Re-infection rate	9.9 %, 8.3 %	6.2 %	>0.05	No difference between SSR and 2-SR
Elshaday et al.16 2020	Meta-analysis of 43 studies	Re-infection rate	6.3 %	10.1 %	0.4	No difference between SSR and 2-SR
Rita et al.17 2023	Meta-Analysis of 44 Studies	Re-infection rate	1.14 %	8.81 %	0.04	SSR group had lesser Re-infection and complication rate than 2-SR group
		Other complications	6.11 %	21.26 %	0.03
Mohamad et al.18 2024	Meta-Analysis of 56 studies	Re-infection rate	10.9 %	12.93 %	0.0062	SSR group had lesser Re-infection and revision rate than 2-SR group
		Revision rate	1.16	2.25	<0.0001

For patients not suitable for multiple procedures, a permanent spacer implantation might be ideal after removal of the infected prosthesis. This might be suitable for very frail patients and who depend on others for their activities of daily life (ADLs). The spacer is impregnated with antibiotics specific to the causative pathogen. This improves the local availability of the antibiotics to the periprosthetic bed as compared to systemic doses. The other advantage is that since the spacer conforms to the space of the removed prosthesis, contracture of the joint is prevented with preserved arm length, better pain management and this in turn leads to better but limited functionality.¹⁹ However, as the tensile strength of the spacer is weak there is a risk of spacer fracture with cement debris accumulation into the joint.

In extreme cases where fitness for surgery is lacking, administration of only suppressive antibiotics may be the only alternative. This is not an effective treatment of choice for eradication because the bacterial biofilm cannot be eliminated in this way, but this may help to relieve some symptoms especially for patients with reduced life expectancy.

6. Conclusion

Prosthetic joint Infection following shoulder arthroplasty is not as common as that in the lower limb counterpart, however, the numbers are increasing, and the post-operative infections are concurrently rising. There seems to be some discrepancies in the causative organisms, sex and age distribution. Propionibacterium acne seems to be the commonest causative organism along with Staphylococcus aureus. This has been attributed to the closeness to the hirsute regions where this organism is mostly resident as compared to enteric pathogen commonly found in the hip infections due to proximity to the lower gastrointestinal system. Management of this condition requires high index of suspicion as this infection is less common and may take a very indolent pathway due to the causative organism. Following the MSIS criteria for suspected infections improves diagnostic accuracy. Treatment options available include radical surgical debridement with thorough irrigation for early cases, 1-stage and 2-stage procedures for late presentation. The option of spacer placement and lifelong use of antibiotics are also available for patients not fit for invasive surgical procedures.

CRediT authorship contribution statement

Saheed Olanrewaju Akinola: Methodology, Investigation, Writing - original draft, Editing. Hariharan Mohan: Writing – review & editing. Zaid Ali: Writing – review & editing. Siddhartha Murhekar: Writing – review & editing. Bijayendra Singh: Conceptualization, Writing – review & editing, Supervision.

Patient consent

Verbal consent obtained from patients for use of photograph and radiographs.

Guardian/Patient's consent

Verbal consents are taken from patients for the photograph and radiographs used in this manuscript.

Ethical clearance

Not required as per local policy.

Ethical statement

Not needed as per local policy.

Funding statement

I do not have any funding disclosure to be made.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.