Pathogens in FRI – Do bugs matter? - An analysis of FRI studies to assess your enemy

Abstract

Fracture-related infection (FRI) is a devasting complication for both patients and their treating Orthopaedic surgeon that can lead to loss of limb function or even amputation. The unique and unpredictable features of FRI make its diagnosis and treatment a significant challenge. It has substantial morbidity and financial implications for patients, their families and healthcare providers. In this article, we perform an in-depth and comprehensive review of FRI through recent and seminal literature to highlight evolving definitions, diagnostic and treatment approaches, focusing on common pathogens such as Staphylococcus aureus, polymicrobial infections and multi-drug-resistant organisms (MDRO). Furthermore, multiple resistance mechanisms and adaptations for microbial survival are discussed, as well as modern evidence-based medical and surgical advancements in treatment strategies in combating FRI.

1. Introduction

Fracture-related infection (FRI), infection in bones or bone fragments following a fracture, is a devastating potential complication after open or closed fractures, treated operatively or non-operatively. It almost exclusively occurs in surgically treated fractures, necessitating a return to the operating theatre for unexpected surgery, prolonged hospital stay and increased healthcare costs. Such infections have troubled the orthopaedic community for decades, multiple definitions and classifications have been used to describe them, with at least nine classifications published in the literature.¹ Until recently, many treatment principles for FRI have been based on the more extensively researched area of prosthetic joint infections (PJI). However, these data and recommendations are not directly transferable to FRI, which has unique features such as the systemic response to injury, bone healing, soft tissue injury, heterogeneity in patient demographics and the patient's general condition. Unfortunately, due to the inherent unpredictable nature of trauma, these factors are seldom amenable to optimisation before surgery.

In December 2016, the first meeting of the Fracture-Related Infection (FRI) Consensus Group took place at the AO headquarters in Davos, Switzerland. This collaboration has brought together members from the European Bone and Joint Infection Society (EBJIS), the Pro-Implant Foundation, the Orthopaedic Trauma Association (OTA) and the AO Foundation with the long-term goal of increasing awareness of FRI to orthopaedic trauma surgeons around the world through publications, websites, and apps. This group has now produced a series of invaluable, practical guidelines based on expert opinion and available research to provide a standardised approach to FRI management for the entire orthopaedic community. These guidelines form a central tenet of this article. However, ultimately, the treatment of FRI is inextricably tied to precipitating microbial infection. This review aims to inform the reader through an analysis of FRI studies and guidelines of an evidence-based approach to assess, diagnose and manage these pathogens.

2. Definitions and diagnosis

2.1. FRI Consensus Group definition of FRI

The various classifications used to describe FRI have lacked consistency and standardisation. Thus, comparing studies and their outcomes has been difficult, hindering the development of uniform, internationally accepted diagnostic and treatment pathways. Furthermore, it has been shown that commonly used time-based FRI classifications may not help predict microbiological epidemiology and, therefore, have limited usefulness in guiding empiric antibiotic therapy.^2,3 In 2018, the FRI Consensus Group agreed upon diagnostic criteria which have since been internationally adopted.⁴ These have been updated to reflect new evidence, including nuclear medicine imaging and histopathological analysis,⁵ and are divided into confirmatory and suggestive criteria, outlined in Table 1. These criteria were subsequently validated with the authors confirming excellent diagnostic discriminatory value.⁶

Table 1.

Diagnostic Criteria for fracture-related infection adapted from Govaert et al.⁸.

Confirmatory Criteria 1. Fistula – sinus – wound breakdown 2. Purulent drainage or the presence of pus 3. Phenotypically indistinguishable pathogens identified by culture from at least two separate deep tissue/implant specimens 4. Presence of microorganisms in deep tissue specimens taken during an operative intervention, confirmed by histopathological examination 5. Presence of more than five polymorphonuclear neutrophils/high-power field, confirmed by histopathological examination

Suggestive Criteria 1. Clinical signs: local – systemic (e.g., redness and fever) 2. Radiological and/or nuclear imaging signs 3. New-onset joint effusion 4. Elevated serum inflammatory markers (ESR – WBC – CRP) 5. Persistent, increasing or new on-set wound drainage 6. Pathogenic organism identified by culture from a single deep tissue/implant specimen

2.2. Diagnosing FRI

As can be seen from the broad nature of these criteria, the diagnosis and confirmation of FRI can be challenging, ranging from the obvious to the obscure. The gold standard for diagnosing an infection is to isolate the bacteria from the wound/fracture site. However, this can be difficult due to evasive microbial adaptations, requiring an in-depth multistep, multidisciplinary diagnostic process.

A major impediment to establishing a widely accepted definition of FRI is that many expected clinical findings in infection are common in routine fracture healing, delayed union or aseptic non-union and are not pathognomonic to FRI.⁷ The updated diagnostic criteria highlight the growing body of evidence about the utility of nuclear imaging, histopathological assessment, and molecular techniques in confirming FRI.^5,8 The findings of these modalities are less likely to be attributed to the natural healing process, unlike elevated non-specific inflammatory markers such as CRP, WCC or ESR levels. This has been formally evaluated, and serum inflammatory markers showed limited diagnostic value in the preoperative diagnosis of FRI, especially when caused by low-virulence microorganisms.⁹

2.3. Diagnosing FRI - imaging

Structural imaging studies such as plain radiography, ultrasound, computed tomography (CT), and magnetic imaging resonance (MRI) are cheap, common and familiar to most medical practitioners. Their efficacy in FRI diagnosis relies on their ability to evaluate fracture consolidation and implant stability and detect abnormal tissue, such as sequestra and bone cavities, or fluid collections that often accompany bacterial infection. However, they lack sensitivity and specificity.¹⁰ Nuclear medicine imaging techniques such as triple-phase bone scan (TPBS), fluorodeoxyglucose positron emission tomography (FDG-PET) and white blood cell (WBC) scintigraphy have the distinct advantage in FRI investigation of detecting biochemical and physiologic abnormalities. These metabolic abnormalities usually precede morphologic changes seen using traditional structural imaging and may be able to identify infection at an earlier stage. Furthermore, complementary hybrid imaging techniques such as single photon emission CT (SPECT), PET/CT, and PET/MRI may help to marry the best of the aforementioned techniques and allow for better anatomic and functional information.¹¹ PET and SPECT are often most helpful in equivocal cases or lesions where tissue sampling is difficult or dangerous.¹² Combining imaging techniques such as WBC scintigraphy with SPECT or FDG-PET with CT may increase the overall sensitivity and specificity in detecting FRI.^10,13,14

2.4. Diagnosing FRI - microbiology

The role of microbiology in diagnosing FRI centres around confirmation of infection, identification of offending microbes and their antimicrobial sensitivities.⁵ Indeed, the culture of phenotypically indistinguishable pathogens from at least two separate deep tissue/implant specimens is a confirmatory criterion in diagnosing FRI. However, despite the promotion of a validated surgical sampling protocol for FRI, false-positive and false-negative culture results can seriously impact patient treatment and prognosis. In addition to this and specific to any infection associated with an implant, the presence of a microbial biofilm must be presumed to be present. Knowing this, it is the clinician's responsibility to ensure tissue or fluid samples are representative and uncontaminated and to discuss the nature of these samples with their microbiologist and microbiology technicians. Consideration should be made of methods to disrupt potential biofilms such as sonication and vortexing with sterile glass beads and culturing samples using appropriate enrichment media for a sufficient duration of up to 10–14 days. Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI–TOF MS) is another advance that allows the identification of bacterial colonies directly from existing media or blood cultures, which is a significant improvement in time, labour and cost compared with prior methods.¹⁵ These efforts can help to maximise obtaining representative cultures and an extended antibiogram of susceptibility to different antimicrobials. Sampling of skin, superficial tissue, or sinus tract samples should be avoided due to their poor ability, even under the best conditions of asepsis, to correlate to intra-op samples, even less so to identify pathogenic/offending organisms.^16,17

Unfortunately, culturing offending pathogens can be difficult as organisms may be fastidious in nature, present only in small numbers, buried within biofilm or in a state of relative quiescence and reduced metabolic activity. Culture-negative FRIs are more frequent in late FRIs (up to 24% of cases) and this may reflect more prolonged antibiotic use in such patients.¹⁸ Molecular diagnostic techniques such as conventional PCR, quantitative PCR, 16s RNA PCR, Sanger sequencing and next-generation sequencing (NGS) involve both targeted or untargeted massive parallel DNA sequencing to detect and amplify bacterial DNA from tiny amounts found in tissue, fluid and sonication samples. NGS has the advantage of being a hypothesis-free approach that does not require prior knowledge of sequence information, unlike qPCR and 16s which can only detect known sequences. NGS therefore provides greater discovery power to detect thousands of target pathogen genes, including rare variants. This may improve sensitivity in confirming the presence of bacteria. They have been shown to identify organisms in over 40% of culture-negative infections assessed initially by traditional methods, that otherwise met diagnostic criteria for infection.¹⁹ However, concerns about the identification of colonising versus actively infectious organisms remain to be resolved as well as the inability of molecular techniques to assess antimicrobial sensitivity and susceptibility. This is a major drawback in the current climate of multi-drug resistant organisms and antimicrobial stewardship. Cichos et al. have described the use of isothermal microcalorimetry as a potential diagnostic tool with considerably decreased time to diagnosis of a fracture-related infection without compromising the accuracy of the diagnosis.²⁰ This tool may allow for early semi-targeted antibiotic management while awaiting speciation and susceptibility testing to modify the antibiotic regimen.

2.5. Diagnosing FRI - histopathological analysis

The use of histopathological analysis of deep tissue specimens in diagnosing prosthetic joint infection (PJI) is well-established and accepted.²¹ Convincing evidence for its use in FRI had been lacking, hence its original omission from diagnostic criteria. However, Morgenstern et al. subsequently showed in chronic/late-onset FRI that the complete absence of PMNs on histopathological analysis had a very high correlation with aseptic non-union (specificity 98%, positive predictive value 98%), while the presence of >5 PMN/HPF was always associated with infection (specificity 100%; positive predictive value 100%).²² This, as well as other evidence, has led to the inclusion of histopathological analysis in the updated criteria.

The discovery and growing understanding of the gut microbiome in relation to general health and illness has led to a rapid expansion in its research.^23,24 Although still very much in an evolutionary phase, its use as an auxiliary diagnostic indicator in suspected FRI patients has already been examined with encouraging results,^25,26 albeit significant work is still needed to corroborate and validate such results.

3. Incidence and prevalence of FRI

The incidence and prevalence of FRI vary worldwide due to the complex interplay of factors such as fracture type (closed vs. open), pathogen, patient demographics, healthcare setting, geographical location and even seasonal variation.²⁷ The average incidence of FRI in surgically managed closed fractures and open fractures is believed to be somewhere between 1 and 5%, and this has remained relatively consistent with time across a diverse cohort of studies.^28,29 A recent study suggests that it may be closer to the lower end of this estimate, with rates as low as 1.23%²⁹ to 1.16%.³⁰ Contrary to this, Walter et al. observed an increase in the prevalence of FRI between 2008 and 2018 in a German population-based study from 8.4 cases per 100,000 to 10.7 cases per 100,000 inhabitants. The most significant increase was in the elderly population.²⁹ It is recognised that the number of fracture-related surgeries has continuously increased, especially in the elderly.³¹ Overall, this may reflect that better initial management of fractures has resulted in fewer new cases developing, but those who are affected, are living longer, but are not cured of infection.

Epidemiological studies suggest an infection rate of 1.8% following closed fractures and 27% following open fractures.^1,32 In cases of severe injury and/or contamination, rates of up to 45% and higher have been reported.^33,34 Wang et al. reported that 65% of FRI patients in their study had sustained an open fracture initially.³⁵ This is in keeping with open fracture being a well-accepted risk factor for FRI. However, there is some data from within the United Kingdom that the rates of open fracture may be decreasing in the last 30 years. Court-Brown et al. previously reported an annual incidence of open fracture as 11.5 per 100,000 individuals,³⁶ while longitudinal data collected by the Trauma Audit and Research Network identifying open fractures reported an overall incidence of 6.94 per 100,000 person-years.³⁷

The precise number of all fractures that require surgical fixation is unknown but it is reported that one-third of all fracture patients are hospitalized with 80% of these admitted requiring surgical intervention.^38,39 Historically, the risk of infection after internal fixation has varied between 0.4% and up to 30% according to the type of fracture (closed or varying degrees of open infection (0.4–16.1%)⁴⁰ Two large contemporary multicentre studies have assessed the risk of infection after internal fixation of closed fractures as 1–2% and of open fractures as 8%.^41,42 Be that as it may, overall fracture-fixation device infections are estimated to comprise less than 5% of all implant-associated infections.⁴³ The vast majority of cases, 75–90%, involve the lower extremity, with the tibia and femur most commonly affected.^1,30 It should be noted that a large number of patients have culture-negative infections, up to 50%,³⁵ possibly due to the administration of antibiotics pre-operatively, poor sampling technique, or not even sampling, so the true incidence could be greater.

The actual incidence and prevalence of fracture-related infection have probably been underestimated due to a lack of precise definition. What is certain is that as the global population continues to grow, the absolute number of patients affected will increase in tandem.

With the advent of an established definition with diagnostic criteria that allows for standardisation of case identification, it may be possible to perform high-quality systematic analysis to calculate a more accurate incidence of FRI.

4. Impact of FRI

The impact of FRI is far-ranging for both patients and healthcare providers. Regarding patient suffering, there is considerable physical, psychological and socioeconomic morbidity associated with FRI. Physical morbidity can be broadly categorised in relation to pain, delayed union and non-union, immobility, loss of function, loss of limb, repeated surgical interventions, and secondary medical issues such as kidney or liver impairment due to prolonged and often high-dose antibiotic therapy. Bezstarosti et al. performed a systematic literature review of treatment and outcomes in fracture-related infection in 2018¹ and identified several stark realities of FRI:

• •Mean infective duration of 28 months, but ranged as high as 154 months
• •Mean age 42 – an age at which many patients are established in the workforce
• •Mean number of operations 3, but ranged as high as 31 operations
• •Mean length of stay in hospital 1.4 months
• •Evidence of non-union in 37%

Unfortunately, to date, the rates of recurrent infection and amputation due to FRI remain high.

Eradication of infection without recurrence varies between 70 and 90% of cases. This can be improved to 93% if repeated treatment protocols, including amputations, are included. Disease recurrence occurs in 6–9% of the patients, and the general amputation rate currently lies between 3 and 5%.^1,44

The psychological burden of FRI on patients, their families and loved ones should not be underestimated and is often overlooked. Only a few studies have specifically looked at the impact of FRI on patients' psychological well-being and quality of life. Bezstarosti et al. included 93 studies on FRI treatment and outcomes in their review, with only three reporting on quality-of-life measures at that time.¹ There is likely a vastly under-reported incidence of mental health-related problems following this life-altering event. Factors influencing this include the timing of diagnosis and the instigation of appropriate treatment. It is even more stressful for patients if the diagnosis is delayed or missed or if there is a failure to begin treatment promptly with the conversion of a simple, early infection into a difficult-to-eradicate chronic infected, painful non-union. Worse physical function and pain scores have been reported in FRI patients compared to non-FRI patients, corresponding to a higher impact on daily life.⁴⁵ Furthermore, FRI suffers have lower quality-adjusted life-year (QALY) profiles than the general population.⁴⁶ Walter et al. showed that even patients who had been successfully treated scored significantly lower on quality of life measures than the reference population, with 21.6% of individuals rated as moderate to severe on the Beck-Depression-Inventory-II (BDI-II).⁴⁴ A lack of information and explanation can make it more difficult for patients to accept the situation, leading to anxiety and other psychological disturbances.⁴⁷ Although not surprising, this highlights the need for formal psychological and patient-peer support for FRI patients to help cope with this experience.

From a socioeconomic point of view, FRI imposes significant, unavoidable consequences for patients and society. FRI is associated with a substantial loss of productivity and potential patient earnings. Considering absenteeism, it can take patients a protracted length of time to return to work after FRI, and a considerable proportion of patients may never rejoin the workforce because of infection and its sequelae.⁴⁵ Beyond this, some patients also contribute to the direct financial cost of their care.⁴⁶ In terms of healthcare economics, the total medical cost of FRI after a tibia fracture is 6.5–8 times that of non-infected patients (additional theatre visits, more complex and longer procedures requiring more advanced and costly treatment options and implants), the length of hospital stay is approximately 8–11 times that of non-infected patients, antibiotic treatment time is 11 times longer, and the cost of consumables (antibiotics, dressing etc.) is approximately 3.6 times higher in infected cases.^45,48 While these data relate to studies performed in Belgium, somewhat similar findings have been reported in the UK,⁴⁶ Spain⁴⁹ and China⁵⁰ and in general can be expected to be extrapolated for the global population.

5. Risk factors

Various risk factors associated with FRI have been identified^1,51, 52, 53 and can be categorised in terms of patient factors, injury factors and surgical factors (Table 2). The following lists are by no means exhaustive, but they do emphasize that most risk factors are essentially non-modifiable being inherent to the patient or the mode of injury, highlighting the importance of patient comorbidities and fracture characteristics in infection development.

Table 2.

Risk factors associated with the development of fracture-related infection (FRI) (MRSA = Methicillin-resistant Staphylococcus aureus, ASA = American Society of Anaesthesiology).

Patient Factors •Advanced/older age •Gender •Diabetes •Obesity/raised BMI •Smoking •Alcohol overuse •Nutritional status •Immunosuppression •Nasal MRSA colonisation at the time of injury •ASA grade 3 or higher

Injury Factors •Lower extremity fracture •Open fractures •Wound contamination •High-energy injuries with bone and soft tissue loss

Surgical Factors •Delayed/prolonged time to surgery •Inadequate debridement with respectful soft tissue handling •Omission or delay of antibiotic prophylaxis •Neglect of dead space management •The presence of implants/fixation hardware •Mechanical instability •Intraoperative hypothermia (<36 °C)

In relation to open fractures the risk of secondary infection according to the Gustilo and Anderson grading is 2–3% for grades I and II, and between 4 and 30% for grade III depending on the severity of soft tissue damage.⁵⁴ Fixation constructs play a direct role in creating a potentially biofilm-friendly environment and in mechanical stability postoperatively. To an extent, this is under the control of the operating surgeon. Fracture instability results in impaired neovascularisation, osteolysis and soft tissue damage, encouraging microbial proliferation, undermining the host immune response and increasing the susceptibility to infection, creating a vicious cycle leading to further biomechanical instability.⁵⁵

Recently, two novel risk calculators have been described for predicting the occurrence of FRI, one for general use in all types of fracture scenarios⁵⁶ and another specific to tibial diaphyseal fractures.⁵⁷ Further study is required to assess their validity in alternative settings but they may help in risk stratification pre-/peri-operatively and allow healthcare personnel to plan surgical intervention and have frank, open discussions with patients and their families regarding the possibility of subsequent infection.

6. Pathogens and pathogenesis

The pathogenesis of FRI involves the establishment and persistence of pathogenic organisms, most commonly bacteria and less commonly fungi, at a fracture site at the time of injury or at a later time point. It may occur in open or closed fractures treated by surgery, with or without the retention of hardware, or even in fractures treated non-operatively. In surgically managed closed fractures, normal skin flora or nosocomial infections predominate, while in open fractures a higher rate of environmental pathogens due to contamination is seen, as well as normal skin flora and nosocomial sources. If the infection is associated with an implant, then biofilm-associated infection must be presumed to be present, making it difficult to eliminate the infection. A large retrospective study identified over 50 different types of bacteria associated with FRI.⁵⁸ The most common single pathogen infections in FRI are Gram-positive cocci (GPC) and Gram-negative bacilli (GNB)⁵⁸ (Table 3). Although the exact percentages vary between regions and are constantly fluctuating, approximately 70–75% are GPC and 20–25% are GNB.³⁵ The microbial profile of FRI most commonly includes Staphylococcus aureus, coagulase-negative Staphylococci (CoNS), Streptococcus species, Enterobacteriaceae, and Pseudomonas aeruginosa.

Table 3.

Microorganisms commonly associated with the development of fracture-related infection (FRI) in descending order of incidence.

Gram Positive (70–75%) Cocci: •Staphylococcus aureus (20–40%) •Coagulase-negative Staphylococci (10–40%) o Staphylococcus epidermidis (5–15%) o Staphylococcus lugdunensis o Non-epidermidis/non-lugdunensis coagulase-negative Staphylococci •Streptococcus species (5–10%) o Streptococcus agalactiae o Streptococcus dysgalactiae o Streptococcus pyogenes o Streptococcus pneumoniae •Enterococcus species (5%) o Enterococcus faecium Bacilli: •Diphtheroid (5%) o Cutibacterium acnes

Gram Negative (20–25%) Bacilli: Enteric (5–20%) •Enterobacter cloacae •Escherichia coli •Klebsiella pneumonia •Proteus mirabilis Non-enteric •Pseudomonas aeruginosa (5–10%) Cocci: •Acinetobacter baumannii Others •Fungi (<5%)

Staphylococci and GNB are responsible for the majority of infections in FRI.⁵⁹ Generally speaking, Staphylococcus aureus (S. aureus) is the most common infection-causing pathogen in FRI overall,^{2,18,35,58,60,61} followed by coagulase-negative Staphylococci such as Staphylococcus epidermidis, Staphylococcus lugdunensis and non-epidermidis/non-lugdunensis coagulase-negative Staphylococci.^2,35,60,61 These are closely followed by Streptococcus species such as Streptococcus pyogenes, Streptococcus pneumoniae, Streptococcus dysgalactiae and Streptococcus agalactiae, with Enterococcus species such as Enterococcus faecium completing GPC pathogens.⁶¹ Finally, Cutibacterium species (Gram-positive bacilli) and Acinetobacter baumannii (Gram-negative cocci) have also been reported as pathogens in FRI.²,⁶¹ In terms of GNB pathogens, Pseudomonas aeruginosa, Escherichia coli, Enterobacter cloacae, Klebsiella pneumonia and Proteus mirabilis are the most commonly encountered bacteria.^2,32,60 Interestingly, it has been reported that bacterial species isolated in FRI do not change significantly over time, with no significant difference in the range of bacterial species isolated in early, delayed or late infections, and S. aureus was the most commonly isolated pathogen regardless of time to onset.^2,18

6.1. Staphylococcus aureus

Staphylococcus aureus is both a commensal bacterium and a human pathogen. It is naturally found in the environment and as part of normal human flora on the skin and mucous membranes. People are the primary reservoir for these organisms, with 30–50% of all adults colonised.^62,63 Higher rates of colonisation (up to 80%) are seen in groups such as healthcare workers, hospitalized patients, and immunocompromised individuals.⁶² This high rate of carriage is the likely driver for S. aureus's predominance as an etiological agent in FRI. It does not normally cause infection in those colonised; however, if it gains entry in the bloodstream or osteoarticular, skin and soft tissue, and pleuropulmonary tissues, this bacterium may cause a variety of serious infections. It uses a number of virulence factors to elude the normal host response including adhesins and biofilm formation, intracellular survival, cytolytic toxins and chemotaxis and complement inhibitory protein production along with production of an antiphagocytic capsule, sequestering of host antibodies or antigen masking by Protein A.^61,62

One such virulence factor is S. aureus's ability to form small-colony variants (SCVs), which are characterized by a smaller colony size, slow and linear growth, altered metabolism and reduced virulence in comparison to wild-type strains.⁶¹ They preside in a quasi-dormant state and grow mostly in protected environments, such as in biofilms and have a penchant to exist within human cells, particularly immune and non-immune musculoskeletal cells.⁶⁴ This adaptation aids to buffer them from extracellular antimicrobial concentration and host immunity resulting in decreased susceptibility to antibiotics and lessened eradication by host immune defences. Furthermore, S. aureus is capable of long-term intracellular infection of non-professional phagocytes, oftentimes called ‘Trojan horse’ macrophages, and bone-resident cell types such as osteoblasts, osteoclasts and osteocytes in vitro and in vivo.^61,65 Finally, S. aureus has been shown to invade the osteocyte lacuno-canalicular network (OLCN) of cortical bone by deforming to approximately half its native size to enter the canaliculi.^61,66 These SCVs are unreachable by host immune cells and explain the persistence and recurrence in clinical cases of S. aureus infection. Once established, this infection may be impossible to clear without radical segmental bone debridement or even amputation.

6.2. Gram-positive cocci

Among the remaining GPC, many are normal inhabitants of the skin, mucous membranes, gastrointestinal and genitourinary tracts and can cause opportunistic infections, especially when medical devices/implants are present. As discussed below, biofilm formation is a common feature of these pathogens. However, compared to S. aureus, they lack the diversity of virulence factors and often present as more indolent infections. This may partly explain the improved efficacy of debridement and antibiotic and implant retention therapy observed in those infections. Additionally, S. agalactiae may preferentially colonise the vascular network within bone as opposed to the OLCN invasion observed in S. aureus infection and may be more susceptible to parenteral antibiotic therapy.⁶¹

6.3. Gram-negative bacilli and others

Enterococcus and the enteric gram-negative bacilli are commonly found in the human gastrointestinal tract and may enter the fracture site through faecal contamination or hematogenous spread. Enterobacter, Pseudomonas aeruginosa, and Acinetobacter baumannii are mainly environmental pathogens that can cause infections in contaminated fractures, with the latter commonly isolated from combat-related injuries.⁵⁹ Pseudomonas aeruginosa is often associated with nosocomial infections. It readily forms biofilms, is notoriously difficult to eradicate, and is known for its intrinsic resistance to many antibiotics.⁶¹ Anaerobic bacteria such as Bacteroides and Peptostreptococcus may be isolated from FRI. These bacteria thrive in environments with reduced oxygen levels. Lastly, fungal infection can be a rare pathogen in FRI. A recent systematic review reported 5 cases of candida FRI. Risk factors included open wounds, prolonged antibiotic therapy, and immunodeficiency. There was a reoperation rate of 33% and the need for protracted antifungal therapy.⁶⁷

6.4. Polymicrobial infection

The previous section relates to monomicrobial infection. Polymicrobial infection rates in FRI are reported as approximately 25–30%.^2,35 Polymicrobial infection is more likely in polytrauma patients, who are usually sicker, immuno-compromised and exposed to nosocomial pathogens. Unsurprisingly, GNB are the predominant pathogens identified in polymicrobial infections, with S. aureus among the most prevalent organisms, along with a variety of co-infecting bacteria. Contamination with skin-commensal bacteria such as CoNS is common post-surgical intervention, whereas contamination in open fractures is more frequently associated with environmental organisms such as E. coli, Enterobacter, P. aeruginosa, A. baumannii, Bacillus, and Enterococcus.⁶¹ Polymicrobial infections have an increased risk of treatment failure due to microbial synergy that enhances virulence and decreases susceptibility to host immune defences.³⁰ It should be noted that standard bacterial cultures often only isolate the dominant organism and so may not reflect the true microbial burden of infection. What's more, many apparent culture-negative infections may actually be polymicrobial, with 90% showing multiple infecting organisms on molecular diagnostic testing. Thus, the high frequency of polymicrobial infections should be considered when choosing an empirical antibiotic regimen for FRI.⁶⁸ To conclude, in cases where no pathogens have been detected despite repeated appropriate sampling, the possibility of fungi and atypical organisms such as mycobacterium should be also considered.

6.5. Multi-drug-resistant infections & organisms

FRIs are potentially difficult to treat when at least one of the disease-causing microorganisms is resistant to biofilm-active antibiotics, including rifampicin-resistant Staphylococci, Enterococci, fluoroquinolone-resistant gram-negative bacteria and fungi.⁶⁰ Worryingly, up to one-third of classifiable organisms causing FRI may be multi-drug-resistant (MDR), even in countries with a low incidence of MDR infection.^59,69 Table 4 lists potential multi-drug-resistant organisms (MDRO) seen in FRI. Geographical differences are seen in the incidence of these infections, but they represent as much as 15–20% of the infective burden.⁵⁸ Nonetheless, a global review of antimicrobial resistance and its future impact estimated that there could be as many as 10 million deaths annually from 2050 caused by resistant bacteria if we do not substantively intervene.⁷⁰

Table 4.

Potential multi-drug-resistant organisms (MDRO) seen in FRI.

Potential multi-drug-resistant organisms (MDROs) include: • Methicillin-Resistant Staphylococcus Aureus (MRSA) • Methicillin-Resistant Staphylococcus Epidermidis (MRSE) • Extended-Spectrum Beta-Lactamase (ESBL)-Producing Enterobacteriaceae • Carbapenem-Resistant Enterobacteriaceae (CRE) • Multidrug-Resistant Pseudomonas aeruginosa (MDRP) • Carbapenem-Resistant Acinetobacter Baumannii (CRAB)

Most microorganisms engaged in material-related contaminations are S. aureus and S. epidermidis, with a significant portion of isolates exhibiting multidrug resistance, posing an immense challenge for clinical practice.⁷¹ MRSA is the most commonly seen MDRO FRI. Over time the number of different strains of MRSA has increased and may reflect the organism's ongoing adaptation to evade clearance. In contrast, the actual rates of MRSA infection have declined steadily in Europe and North America,⁷² most likely related to improved antibiotic stewardship, education, improved hand hygiene and perioperative decolonisation. Despite this, the overall rates of MDRs in general have remained relatively stable at 15–17%.⁶⁹

6.6. Microbial resistance mechanisms

Similar to other organisms, bacteria evolve through a process of heritable change in populations over multiple generations. This occurs via natural selection and genetic drift allowing for new genetic and physical adaptations to develop that help bacteria to adjust to environmental change or host immunity. However, this also makes bacteria susceptible to environmental pressures under stressed conditions. In response to this, with their ability to reproduce at extremely fast rates with short generation times and large population sizes, bacteria can rapidly produce new strains with beneficial survival changes.

Multiple resistance mechanisms and adaptations for microbial survival have been described and can generally considered as intrinsic, acquired, and adaptive.⁷³ Intrinsic resistance pertains to the natural features or characteristics of a bacterium and its innate properties that limit the action of antibiotics. Of particular interest to FRI is the property of biofilm formation, which will be discussed below. Genetic mechanisms can be associated with acquired or adaptive changes in antibiotic sensitivity. Traditional acquired antimicrobial resistance centres around de novo mutations during replication or horizontal transfer of genes via the acquisition of plasmids and mobile genetic elements. Intrinsic and acquired resistance are stable, being transmitted vertically to subsequent generations. In contrast, adaptive resistance can modulate gene expression, possibly by epigenetic alterations, and is often unstable, reverting following the removal of the stressor. These mechanisms include hypermutability and small colony variants (discussed above), as well as persister cells, adaptive resistance, and Phoenix colonies. A detailed account of these is beyond the scope of this paper and can be found elsewhere.^64,73

6.7. Biofilms in FRI

According to the National Institutes of Health (NIH) and the 2018 International Consensus Meeting on Musculoskeletal Infection approximately 60–80% of all microbial infections are linked to biofilm formation.^74,75 What is more, Zimmerli and Sendi have previously highlighted the role of biofilms in FRI and antibiotic resistance.⁷⁶ A biofilm has been defined as a microbe-derived sessile community characterized by organisms that are attached to a substratum, interface, or each other, are embedded in a matrix of extracellular polymeric substance, and exhibit an altered phenotype with respect to growth, gene expression, and protein production.⁷⁴ It is a complex and well-structured three-dimensional aggregation of single or multiple microorganism species, often with a dominant organism orchestrating the infection. It is a bacterial community transitioning from a free-floating individual planktonic state to a sessile lifestyle. It operates through quorum-sensing and cell-to-cell signalling, allowing individual bacteria within colonies to orchestrate colony-wide functions via signalling molecules called autoinducers, controlling gene expression that dictates collective behaviours.

Biofilms are composed of a cellular component of bacterial cells (∼10%) and a noncellular component, also known as a glycocalyx, derived from water, polysaccharides, nucleic acids, proteins, lipids, and lipoproteins (∼90%). The noncellular components can vary greatly with the type of bacteria.⁷⁷ Biofilms can either be single or multi-layered, with various micro-niches occupying these layers. Their formation begins as soon as planktonic bacteria encounter a surface, ideally an abiotic surface such as devitalised bone or foreign/prosthetic material. Rough surfaces are more susceptible to biofilm formation due to reduced shear forces and increased surface area for colonisation. In relation to FRI, the non-motile bacteria S. aureus and S. epidermidis are prime examples of this process. Initially, attachment is reversible but soon becomes irreversible. Several factors contribute to this adherence, including hydrophilic and ionic interactions and surface adhesion proteins such as microbial surface components recognising adhesive matrix molecules (MSCRAMMs) and secretable expanded repertoire adhesive molecules (SERAMs).⁷⁸ Following this, cell-to-cell adhesions creating aggregations of bacteria begin to secrete extracellular matrix to form a monolayer and subsequently multi-layered microcolonies with the initiation of quorum sensing. Finally, these biofilms mature into large microcolonies, which can undergo detachment and dispersal to set up a cycle of persistent infection.

A variety of factors have been linked to antibiotic resistance in biofilm infections. The physical barrier offered by the layers of material within a biofilm essentially provides a protective blockade for colonies from host defences while also limiting diffusion and decreasing the penetration of antibiotics.^79,80 Travelling towards the centre of a biofilm, bacteria exhibit altered metabolic activity and growth kinetics as access to nutrients and oxygen is limited and waste products accumulate. Cells phenotypically change from active planktonic bacteria to metabolically dormant “persister” cells, entering a slow-growing or non-growing state with the ability to persist in the G0 phase of the cell cycle.⁶⁴ Therefore, antimicrobial agents that target bacteria during the growth phase or interfere with metabolic activity are rendered ineffective. Persister cells are seen in S. aureus populations⁶¹ and can be up to 1000 times more resistant to most antimicrobial agents than their planktonic relatives.⁸¹ Once the antimicrobial threat is removed or reduced below therapeutic levels, persister cells can re-enter the growth cycle to re-populate the biofilm, causing prolonged infections. S. aureus can also excrete several proteins that make it less vulnerable when inside a biofilm, such as the Staphylococcal complement inhibitor (SCIN), the chemotaxis inhibitory protein of Staphylococci (CHIPS), the clumping factor A (ClfA), and the extracellular adherence protein.⁸² Another proposed mechanism in the literature for acquired drug resistance in biofilms is horizontal gene transfer by conjugation, transformation, transduction and outer membrane vesicle-mediated transfer of genetic material due to the close proximity/stabilised cell-to-cell contact and long retention time of cells within biofilms.⁸³

As a final point, most units handling large volumes of FRI are tertiary referral centres, often dealing with cases that have been transferred for expert management as they have failed treatment elsewhere. Hence, there may be an unintentional selection bias towards more difficult-to-treat organisms in this population of patients.

7. Management

The goals of management of FRI are the eradication of infection, fracture union and restoration of function to the affected limb/area. This typically involves both medical and surgical approaches (see Table 5 for potential strategies in the management of FRI) and considers factors including microbe virulence, patient health (such as underlying medical conditions), limb health (such as fracture configuration, soft tissue component to the injury, contamination), and the treating surgeon's experience including available resources. Ideally, treatment should only be started after an appropriate clinical assessment and diagnostic workup is complete, focusing on the elements of the Consensus Definition of Fracture-related Infection. Senior clinicians should review patients throughout, and each case should be considered individually, with initial assessment guiding the urgency of care. Early infections may be managed in the acute trauma unit with microbiological support or referred onwards. Suspected late fracture-related infections (e.g. infected non-unions, infected healed fractures) in a well-patient should be seen at a specialist bone infection combined clinic.⁸⁴

Table 5.

Potential surgical strategies in the management of FRI.

Patient not suitable for surgery • “Dress and suppress” Patient suitable for surgery and fracture united • Debridement, removal of metalwork±local antibiotic therapy Patient suitable for surgery and fracture not united (single stage surgery) • Debridement with retention of hardware (DAIR) if fixation is stable, or if reasonable alternative fixation is not possible or practicable±local antibiotic therapy • Debridement with removal of hardware and exchange of metal work or alternative stabilization±local antibiotic therapy Patient suitable for surgery and fracture not united (two stage surgery) • Debridement, temporary removal of all metal work±spacer ± alternative fixation (e.g., external fixation) ± local antibiotic therapy, with later reimplantation of fixation Patient not suitable for recurrent surgeries/major reconstructive surgery, overwhelming sepsis due to FRI or patient wishes • Amputation

Note: These strategies are used in tandem with systemic antibiotic therapy.

7.1. FRI multidisciplinary team

Effective protocols for infection prevention, detection and management must be implemented as part of a multidisciplinary service. All medical management should be consultant-led, and all patients with proven or suspected FRI should be the subject of regular bone and joint infection MDT review to rationalise surgical strategy and antimicrobial management.84, 85, 86 This MDT should include orthopaedic surgeons, plastic surgeons, microbiologists and infectious disease physicians/pharmacists, musculoskeletal radiologists, specialist nurse practitioners, rehabilitation specialists and administrative support.84, 85, 86 Nursing and secretarial team members often provide a critical link between patients and medical personnel, ensuring patients are well-informed. This simple intervention is known to improve the experience of patients with FRI.⁴⁷ Within the author's unit, our rehabilitation team facilitate a “frame club,” allowing patient-peer support to exchange experiences. Ideally, a team member should be trained in providing psychological support, as patients can readily develop anxiety and other psychological disturbances.⁴⁷

7.2. Prevention and perioperative management of FRI

As per BOAST guidance, a primary goal in fracture management should be preventing infection.⁸⁵ To this end, the reader can refer to in-depth review articles focusing on pre-, intra- and post-operative practical strategies and advice on infection prevention.^52,87 In terms of infection prevention, systemic antibiotics administered before surgical incision is one of the most effective evidence-based preventive therapies in implant-associated surgery.⁸⁷ In terms of infection management, antibiotic therapy is typically used in conjunction with surgery to control and clear infection. Generally, there is no role for empiric antibiotics in suspected FRI without a diagnostic workup, especially in a well-patient, and they should be stopped if already commenced.^84,85 However, any patient with signs of systemic sepsis should have an immediate medical assessment with blood cultures taken before the urgent commencement of parenteral antibiotic treatment.^84,86 In suspected cases of FRI in a patient who is not systemically unwell, a review by a consultant in a clinic within 48 hours should be arranged. Antibiotic treatment should only be commenced after that review and can be safely withheld until reviewed or at least discussed with the treating fracture surgeon. Importantly, in well-patients, co-morbidities should be addressed before definitive infection treatment but should not produce indefinite delays.^84,86 Optimisation of peri-operative smoking cessation, nutritional and vitamin D status, as well as glycaemic control can have a beneficial impact on outcomes.⁸⁶

7.3. The role of antibiotics in FRI

On occasion, patients may be definitively treated with antibiotic therapy alone, particularly in older patients or in known longstanding FRIs who have already had multiple previous operations, have poor soft tissues and where the risks of definitive surgery may be unacceptably high to the patient or the treating surgeon, compared to on-going disease symptoms.^86,88 In these patients, a diagnostic workup may be considered, followed by either no therapy or antibiotic suppressive therapy until fracture consolidation (“dress and suppress”). If a suppressive approach is taken, antimicrobial agent choice should be based on culture sensitivities, have a low rate of resistance induction, be orally available and have an acceptable side effect profile for prolonged use.⁸⁹ In chronic infections, patients may have long periods of disease quiescence requiring no treatment, and flair-ups may be treated based on previous sample results and clinical status. However, the overuse, misuse and ineffective use of antibiotics should be avoided to prevent the development of antibiotic resistance. Therefore, in keeping with guidelines on good antibiotic practice and stewardship,^84,85 each event should be evaluated on a case-by-case basis, and all antimicrobial plans should be made in conjunction with an infection specialist.

Generally, once FRI has been confirmed, ideally after appropriate microbiological sampling, empiric broad-spectrum antibiotic therapy should be commenced systemically at the first opportunity. During the early stages of infection, bacteria are in a planktonic phase, rapidly dividing with short generational cycles and thus susceptible to systemic concentrations of antibiotics. A delay in commencing systemic antibiotics may allow for mature biofilm formation. This is a significant therapeutic challenge as the high systemic antimicrobial concentrations required for biofilm suppression and eradication may cause acute bone marrow suppression and renal and hepatic toxicity.⁸⁶ Hospitals should have a relevant antibiotic policy for the initial management of suspected infected metalwork after sampling that is based on geographical and temporal differences of likely pathogens, local antibiograms, antimicrobial susceptibilities/resistance, patient co-morbidities and allergies, disease severity, and injury history.^85,86 In all cases, this should be adjusted based on sampling results.⁸⁵ This policy should cover commonly offending pathogens such as Staphylococci and GNB, which are responsible for most FRI. Furthermore, as these patients will have previous hospitalisations for their fixation surgery, clinicians must remain vigilant to the possibility of polymicrobial, nosocomial and MDRO infections.^2,59

Several potential antimicrobial combinations to achieve initial adequate coverage and later definitive treatment have been proposed including the use of a glycopeptide (usually vancomycin or teicoplanin) with at least one of the following: gentamicin, co-amoxiclav, a carbapenem (meropenem or ertapenem), a cephalosporin (such as ceftriaxone) or piperacillin/tazobactam.^60,90 Of these agents, gentamicin and vancomycin are also routinely used for local antibiotic delivery in dead space management, discussed subsequently. Definitive antibiotic treatment is tailored towards sample results and is based on culture sensitivities, patient co-morbidities, allergies and disease severity. Recently published consensus guidelines from an international expert group provide an excellent in-depth review of potential targeted antimicrobial treatment based on specific infecting pathogens and offer standardised recommendations for antimicrobial therapy in terms of agent of choice, dose, frequency, administration route and duration.⁸⁹ Furthermore, good outcomes in terms of eradication of infection with limb preservation in FRI patients have been published based on these guidelines.⁹¹

Traditionally, antimicrobial treatment has involved prolonged courses (6–12 weeks) of intravenous antibiotics, associated with substantial risks, inconvenience, and higher costs than oral therapy.⁹⁰ Out-patient parenteral antibiotic therapy (OPAT) services may help in allowing patients home therapy in situations where oral antibiotics are not an option. However, Li et al. have shown in the OVIVA trial that the use of oral antibiotic therapy, after 1 week of empiric broad-spectrum IV antibiotics, was not inferior to intravenous antibiotic therapy when used during the first 6 weeks for complex bone and joint infections.⁹⁰ This has prompted a change in practice promoting the early transition to oral antibiotic therapy in combination with antibiofilm antibiotics, such as rifampicin or quinolones. Crucially, the prerequisite to this approach is achieving local infection control first.

7.4. Surgical management of FRI

The propensity of commonly infecting bacteria to produce biofilms can render isolated antibiotic therapy ineffective in eradicating infection. Empirical antibiotics alone have shown unfavourable outcomes with the risk of chronic infection, antibiotic-related side effects, and the development of bacterial resistance.⁹² Therefore, local control by debridement, implant removal or exchange, and in severe cases, amputation are critical to reduce the bacterial load at the fracture site and increase the efficacy of antimicrobials.⁸⁹ Currently, the two main surgical approaches to FRI management are single-stage debridement, antimicrobial therapy, and implant retention (DAIR) or debridement, antimicrobial therapy and implant removal/exchange, which may be a two-stage process. Critically, common to both is an adequate debridement. The emphasis must be placed on adequate debridement, even though it may increase the complexity of reconstruction. Anecdotally, a common error of many orthopaedic trauma surgeons is the failure to excise enough compromised bone or soft tissue due to concerns about the ability to reconstruct the bone defect or soft tissue envelope. Therefore, the role of the MDT, prior discussion with plastic surgeon colleagues and dual surgeon operating may result in optimal debridement.⁸⁶ To highlight the importance of this, there has been a change in terminology used in the 2020 edition of the “Standards for the Management of Open Fractures” from the British Association of Plastic, Reconstructive and Aesthetic Surgeons (BAPRAS) and British Orthopaedic Association (BOA) from debridement to wound excision to promote the concept that completion of the process relies on more than just lavage and dilution of contaminants.⁹³

7.4.1. The importance of debridement in FRI

Although general treatment principles for FRI have been published,⁸⁶ important principles on debridement techniques can be extrapolated from the extensive research base of PJI.94, 95, 96 These can be further supplemented with guidance from the Arbeitsgemeinschaft für Osteosynthesefragen (AO) Foundation. Debridement should be a meticulous, organised process to maximise the excision of infected material and minimise bacterial bioburden. It can be categorised into surgical debridement, mechanical/physical debridement and chemical debridement. A central tenet of debridement is the taking of multiple deep, uncontaminated tissue samples for bacterial and fungal cultures.

Surgical debridement should follow an orderly progression through tissue layers, starting superficially and working into the depths of the wound. All non-viable or questionable skin should be excised, including any sinus tracts. The incision can be extended as needed, preferably along perforator-sparing fasciotomy lines, for wide exposure of the pathological zone. Skin and fascial extensions should be raised as a single unit in the sub-fascial plane to ensure they constitute a robust flap.⁹³ Any exudate such as pus, seroma or hematoma, abscess membrane, fibro-inflammatory tissues or granulation tissue should be removed, and any extension of suspicious fluid or tissue beyond the original surgical field should be thoroughly evaluated. Any necrotic soft tissue (fat, muscle, fascia) should be excised. Neurovascular structures and viable tendons must be identified and preserved. Once the depths of the wound are exposed, debridement includes excision of all necrotic, non-viable and questionable bone, along with any poorly perfused tissue and the removal of all non-essential foreign material, especially loose hardware/fixation. Bone resection should include sequestrectomy and continue back to viable bone indicated by small punctate bleeding osseous vessels commonly called the “paprika sign”. In cases of FRI with intramedullary fixation, instruments such as the Reamer-Irrigator-Aspirator (RIA) System® (DePuy Synthes, Leeds, England) facilitate intramedullary and endosteal debridement.

It must be stressed that local and systemic antibiotics are not a substitute for proper surgical technique, respectful soft tissue handling and thorough debridement. Adequate debridement can take time, should not be rushed and should result in a clearly defined margin of healthy tissue removing any infected material or hardware that could provide an avascular substrate for biofilm formation. Well-established infections pose a challenge as there may be no clear demarcation between healthy and infected tissues. Methylene blue stain has been described as a promising technique to intra-operatively preferentially highlight biofilm-infected surfaces,⁹⁷ which can then be removed, although further investigation is needed in FRI. Despite this, in extensive disease and infective burden or if adequate debridement is impossible, radical excision, including amputation may be necessary to eliminate infection.

Mechanical/physical debridement using a sterile scrub brush with or without 2% Chlorhexidine gluconate to mechanically disrupt and remove bacteria and biofilm from the surfaces has been described in PJI and may be used in FRI.⁹⁸ Additional techniques described include non-contact induction heating of metal prostheses, pulsed electromagnetic fields, low-intensity pulsed ultrasound, and thermically-guided removal,^94,99 although their reported use in FRI is sparse. After debridement, the surgical site should be irrigated with copious amounts of normal saline solution to decrease bacterial load, remove loose debris and facilitate visual examination of the wound. The International Consensus Meeting group has given a recommendation for the use of 6–9L of normal saline irrigation in PJI.⁹⁶ Current guidelines for FRI also advocate using normal saline but at low pressure. This is based partly on the landmark Fluid Lavage of Open Wounds (FLOW) trial,¹⁰⁰ wherein investigators found no difference between high or low-pressure irrigation with sterile saline in open fracture care and the belief that high-pressure lavage may drive contamination deeper into tissues. However, the use of pulsatile lavage may have a place in the management of an established infection but there is currently no literature that supports the use of one method over the other. Irrespective of this, irrigation should be continued until the wound is macroscopically clean.

To date, the optimal irrigation fluid is still unknown and remains an area of controversy. The 2015 FLOW study only assessed saline vs. soap and several alternative irrigants have been proposed since then. These include topical antiseptic agents such as povidone-iodine, chlorhexidine gluconate, the weak organic acid acetic acid alone or in solution with benzalkonium chloride (Bactisure, Zimmer–Biomet, Warsaw, IN, USA), hydrogen peroxide, polyhexanide, ethylenediaminetetraacetic acid and octenidine.^94,101, 102, 103, 104, 105, 106 If alternative irrigation solutions are used, normal saline lavage should be performed between solutions and as the final solution. It is clear that further research is required on the optimal irrigation protocol. Finally, surgeons may choose to have all drapes, even gowns, gloves, and instruments exchanged for clean ones following debridement and irrigation. Although the advantage of the theoretically decreased bacterial burden of the surroundings of the surgical site has not been definitively proven, this simple, cheap and short practical step to divide the procedure into “dirty” and “clean” stages would intuitively seem beneficial. Indeed, in cases with significant amounts of contaminated, necrotic dead tissue, or grossly purulent wounds, a “second look” debridement can be repeated to ensure surgical clearance of infected tissue back to a healthy base has been achieved.

7.4.2. Single vs. two-stage management of metalwork

As mentioned earlier, the two main surgical approaches to FRI management are single-stage debridement, antimicrobial therapy, and implant retention (DAIR) or debridement, antimicrobial therapy and implant removal/exchange, which may be a two-stage process. Divided literature and heterogeneous data make it difficult to draw definitive conclusions on implant retention or removal during treatment of FRI. However, fundamental to the decision-making process is the integrity of the fixation and whether or not the fracture has consolidated, with hardware maintained if stability is at risk with its removal. The DAIR concept was originally described in PJI¹⁰⁷ and many lessons have been learnt from its evolution. It is a potentially less morbid intervention than the formal revision of fixation, but results may vary. It can be considered for acute/early FRI, with a short duration of infection (up to three weeks) and under strict preconditions, such as proven construct integrity and fracture stability, as well as identifying a sensitive organism with an appropriate antimicrobial agent available. In these circumstances, successful retention rates may be as high as 86%–100%.^30,108 Chronic/late onset FRI may be associated with high rates of recurrence if treated with DAIR.¹⁰⁸ Because of this time-dependent aspect, DAIR should be considered an urgent, but not emergent procedure, allowing time for a thorough workup in well-patients.

In cases with unstable implant fixation, unacceptable fracture reduction, poor host physiology with severe or extensive infection, implant removal or exchange must be considered. Furthermore, certain pathogens, such as P. aeruginosa, multi-drug-resistant organisms, such as MRSA, polymicrobial infections, and fungal organisms are associated with higher failure rates with DAIR and should be considered for alternative treatment.^30,109 Whether internal fixation is exchanged or completely removed depends on several factors but primarily on fracture consolidation. If the fracture is united or if exchange of internal fixation is not possible due to bone loss or poor soft tissue coverage, then internal fixation may be completely removed. In the second scenario, re-stabilization may be performed with external fixation. This versatile construct can allow ongoing management, ensuring that transfixion wires and half pinas are positioned away from the infective bed, while still maintaining adequate stability. Exchange protocols may involve one or 2 stages. In either case and following DAIR, the use of an antibiotic-loaded carrier placed in or around a fracture during the surgical treatment of FRI is likely to reduce the risk of recurrence of infection and prevent recolonization of metalwork. This may be in the form of bioabsorbable ceramics^110,111 or antibiotic cement-coated plates.^112,113

7.5. Dead space management in FRI

Adequate debridement in FRI often creates a bone and soft tissue defect. Without timely obliteration of this dead space, thereby preventing hematoma and seroma formation, and the replacement of bone and soft tissues with viable vascularized tissue capable of distributing systemic antibiotics, this poorly perfused area can potentiate bacterial proliferation and persistence. A number of strategies exist in dead space management ranging from temporary antibiotic-eluting bone void fillers, prosthetic replacement, induced membrane and grafting techniques, to bone transport and distraction osteogenesis, even amputation, as well as skin grafts with local and free flaps for soft tissue coverage. The list of strategies continues to expand and what follows is by no means an exhaustive review.

7.5.1. Local antibiotic delivery in FRI

Consensus guidelines on the treatment of FRI have advocated that the use of local antimicrobials needs to be strongly considered.⁸⁶ This was further supported by a recent clinical study investigating FRI management, which has suggested that high local antimicrobial concentrations can eradicate/exhaust most antimicrobial variants, regardless of susceptibility. Gentamicin, tobramycin and vancomycin are commonly incorporated into the production of beads, pellets, bone void fillers and spacers used for both prevention and treatment of FRI. Other antimicrobials can also be used, including daptomycin, amikacin, and voriconazole.¹¹⁴ These medicines must be sterile, heat-stable, water-soluble, and in powder or liquid form with a reasonable volume, while also being broad-spectrum with a low risk of resistance induction, hypersensitivity and/or allergies.¹¹⁵ Local antibiotic delivery vehicles are made from materials including polymethyl methacrylate (PMMA) bone cement, collagen sponges, biodegradable polymers and hydrogels, and calcium-based biocomposites (ceramics and bioglass).¹¹⁴ These may be used in their own right or as part of commercially available fixation devices such as the PROtect tibia nail (DePuy Synthes; Johnson/Johnson Company, Inc, New Brunswick, NJ).

Originally the addition of antibiotics was an ad hoc, off-label use with fears relating to impaired mechanical properties, poor elution kinetics, and local and systemic toxicity but is now a well-accepted and trusted practice. Indeed, in large segmental defects, spacers can be reinforced and fashioned to provide a degree of structural integrity allowing for at least partial weight bearing in combination with internal or external fixation. Furthermore, the reported antibiotic elution profiles of these delivery vehicles have been extensively studied and have been shown to be capable of delivering antibiotics above the minimum inhibitory concentration (MIC) for several weeks after implantation, particularly the calcium-based biocomposites with superior elution profiles to PMMA.¹¹⁶ That being said, concerns have been raised regarding the effect that antibiotics have on cell viability, cellular toxicity, osteogenic potential and fracture healing. Reassuringly, it has been demonstrated using preclinical in vivo models and in clinical studies that local application of gentamicin±vancomycin, in standard doses, does not appear to adversely interfere with fracture healing.^117,118 Furthermore, local gentamicin application has been shown to be effective against both methicillin-sensitive and methicillin-resistant S. aureus in vivo and in combination with oral rifampicin has been reported to reduce the minimum biofilm eradication concentration (MBEC) needed to produce this effect in a preclinical model.¹¹⁹ Amikacin, tobramycin, and vancomycin have also previously been shown to produce low cytotoxicity until very high concentrations were used.¹²⁰

PMMA has historically been the most popular strategy for dead space management and delivery vehicle for local antibiotic therapy in PJI and FRI, with more than 25 years of published use in infection.¹²¹ It is cheap, readily available, easy to use and versatile allowing for the fabrication of beads and custom-shaped spacers to be formed to conform to the infective bed. Antibiotics and antifungals may be added to it during mixing, but several companies now supply premixed, antibiotic-loaded cement. PMMA is commonly used as a spacer in reconstructive techniques such as the Masqulet/induced membrane technique. Masqulet has reported success ranging from 75% to 88% for infection clearance and osseous reconstruction.^17,122,123 In addition to this, if segmental excision and an appropriate debridement are performed, 2 weeks of antimicrobial therapy may be all that is needed to eradicate any residual contamination in the soft tissues.⁸⁶ Furthermore, this 2 stage technique allows the pairing of bone grafts with tailored antibiotics based on the initial debridement. Although, this induced membrane theoretically provides an ideal biological scaffold for dead space management and infection control, with osteogenic cells and factors but without adverse effects on graft incorporation,¹²⁴ further large-scale studies are needed to provide robust evidence.

PMMA's use in FRI management also includes its off-label coating of temporary and definitive internal implants, including plate and IM nails. Despite its obvious benefits and real-world clinical use, some issues regarding PMMA's use as an antibiotic-loaded carrier have been discussed elsewhere.^116,125 Different elution rates and concentrations have been noted¹²⁶ and should be considered when choosing bone cement. The primary release of antibiotics is in a burst fashion in the first 48–72 hours after implantation. However, ongoing elution may be less than the MIC needed to effectively kill organisms and may select out resistant organisms, even allowing the PMMA to become colonised itself.^114,116 Furthermore, antibiotic elution from PMMA is via diffusion and this process is partly determined by the antibiotic itself, its proportion in the cement, surface area, porosity of the cement and possibly even the composition of the cement. Increasing porosity improves antibiotic escape and elution but may impact its mechanical properties, putting it at risk of fragmentation and debonding on coated plates, nails or cement rods.112, 113, 114 Finally, as PMMA is not biodegradable, it requires a secondary procedure for its removal.

In comparison, antibiotic-containing collagen sponges, hydrogels, and calcium-based biocomposites (ceramics and bioglass) are dissolvable and so all of the antibiotics contained within them will be released resulting in high local antibiotic concentration gradients and consistent and persistent degradation patterns without the prolonged low-level release of antibiotics observed with PMMA.¹¹⁶ In addition, a calcium sulphate hydroxyapatite biocomposite, Cerament® G (with gentamicin) or V (with vancomycin) (Bonesupport, Lund, Sweden) has been shown to produce radiographic resolution of bone defects in the treatment of osteomyelitis and was associated with osseous tissue formation, as supported by histological analysis.¹²⁷ Additionally, this flowable, cold-curing composite can be injected into the IM canal in a fluid state before nail insertion. The carrier therefore coats the surface of the nail, potentially protecting it from colonisation and delivering a high local dose of antibiotic.¹¹⁴ A similar protective coating of implants has also been described for antibiotic-loaded hydrogels.¹¹⁸ Alternatively, these adjuncts can be placed into former screw holes, intraosseous defects or intramedullary during DAIR procedures or at the first stage of a two-stage exchange protocol. However, these strategies, although generally considered safe, are not without potential problems. Systemic toxicity and iatrogenic hypercalcemia have been reported in patients undergoing calcium-based antibiotic carrier implantation for infection, albeit mainly for PJI.^128,129 Furthermore, prolonged wound leakage and delayed wound healing have also been reported.^116,130,131 Thankfully, the overall numbers of such complications are low and generally do not lead to long-term problems.

An important part of the surgical management of FRI is the soft tissue reconstruction frequently required. An in-depth review of this topic is beyond the scope of this paper but suffice to say that the optimal treatment depends both on the surgeon's expertise and ability and specific patient factors. However, early closure with a healthy, vascularized soft tissue envelope is an essential prerequisite. It should also be stressed that negative pressure wound therapy is a temporary adjunct towards closure.¹³² Finally, amputation must always be considered as an appropriate form of treatment and not viewed as a sign of failure. It may be a lifesaving procedure in some patients and indeed a not insignificant number of patients present or develop acute sepsis requiring emergent care due to FRI.³⁰ In addition to these, some patients may have unreconstructible defects or choose amputation as a reliable, trusted treatment option instead of facing a prolonged, somewhat unpredictable limb reconstruction process.

8. Future treatment options

Due to a number of reasons including the financial cost of development, preclinical and clinical trials and regulatory approval, there has been limited on-going development and production of new antibiotics.¹³³ To highlight this, in 2021 the WHO identified only 27 new antibiotics in clinical development against priority pathogens, down from 31 products in 2017 and only six of these were classified as innovative.¹³⁴ This has focused attention on developing strategic blueprints to substantially improve the discovery and development of new antibiotics.¹³⁵ Aside from this, alternative therapeutic agents and avenues of treatment are being explored to tackle this therapeutic niche not well covered by antibiotics, some of which are new and some of which are re-explorations of previous interventions. Some of these strategies are broadly outlined in Table 6. Due to biofilms' commonness and clinical implications, much of the research into device-related infection has centred on this area. Excellent reviews of potential strategies have been published and provide an informative overview of this topic.136, 137, 138

Table 6.

Potential novel strategies for the management of FRI.

• Biologics o Polyclonal and monoclonal antibodies, antimicrobial peptides • Endogenous molecules/viruses that target bacteria including bacteriophage therapy • Photodynamic inactivation (PDI) • Surface coatings and modification to prevent recolonization • Specific targeting of biofilm components through enzymatic and endogenous disassembly molecule activity promotion • Continuous local antibiotic perfusion (CLAP) • Small molecule inhibition, nanoparticles and quorum quenching agents o Via cyclic nucleotide second messenger in bacteria signalling pathway and nitroxides • Electrochemical methods o Cathodic-voltage‐controlled electrical stimulation (CVCEs), electrochemical scaffolds and hyperthermia

As discussed earlier, intracellular bacteria can be a difficult challenge in FRI eradication. Monoclonal antibodies that target matrix components of biofilms can act as both therapeutic and diagnostic molecules through the delivery of conjugated antibiotics, radionucleotides, biofilm-degrading enzymes and photosensitizers. Once bound to bacterial cell surfaces, they undergo opsonization, degradation and release of their payload within the bacterial cell, including antibiotics.¹³⁹ This pathway, as well as the addition of cell-penetrating peptides or liposome nanocarriers to antimicrobials, offer a promise of true eradication of persistent infection.

Bacteriophages have been in clinical use since the First World War to treat bacterial infections, even before the advent of antibiotics. Due to antibiotic resistance, there has been a recent resurgence of interest in phage therapy as it is not affected by many of the traditional antimicrobial resistance mechanisms or bacterial cell dormancy. It utilizes bacteriophages, viruses that are highly specific in their ability to target particular strains or species of bacteria, to infect and replicate within these bacteria, ultimately causing cell lysis and destruction.¹⁴⁰ These bacteriophages can move through the population of bacteria within biofilms and significantly reduce viable numbers of cells. This mechanism of action allows phages to specifically target pathogens while leaving potentially beneficial and normal flora.¹⁴⁰ Although specific to target pathogens, it has the potential to be broad spectrum and target a wide range of bacterial species, including antibiotic-resistant strains. Additionally, they may provide an alternative or adjunct to antibiotics, theoretically reducing the reliance on antibiotics. Phage therapy is potentially cheap with the ability to produce large-scale phage products, as already established commercially in the food industry.¹⁴¹ However, bacteria can develop resistance to phages through changes in the membrane-based targets including lipopolysaccharide, preventing phage adsorption. It is important to note that while phage therapy shows promise, it is still considered experimental and should be limited to the treatment of infections after intolerance or ineffectiveness of antibiotic therapy.¹⁴¹ Despite the challenges and considerations, including safety, efficacy and regulatory considerations, appropriate delivery methods, and the need for well-designed clinical studies,¹⁴² bacteriophages are potentially an attractive therapeutic agent.

Another novel approach to FRI treatment in terms of biofilm destruction is the use of photodynamic inactivation (PDI). This uses light in combination with a photosensitizer molecule and molecular oxygen to generate reactive oxygen species which produce oxidative stress eventually leading to microbial cell death.¹⁴³ PDI is relatively cheap and non-toxic and can be used on implants in situ. Furthermore, it has been shown to be effective at killing nosocomial pathogens within biofilms in vitro or significantly reducing the number of viable cells with clinically realistic illumination times,¹⁴⁴ including clinical evidence to support its use in wounds infected with P. aeruginosa, MRSA and A. baumannii in patient¹⁴⁵ and mouse models.146, 147, 148 Interest has also grown in the design and development of surface modification of medical implants, in part due to the discovery of the natural bactericidal properties of cicada wings due to their nanopillar arrays causing mechanical rupture of Pseudomonas aeruginosa cells. This has expanded over time, with interest in graphene-based carbon nanomaterials and metal-based nanomaterials to exhibit strong antimicrobial and anti-adhesive activities that can inhibit microbial colonisation of clinically relevant bacteria such as S. aureus and E. coli.149, 150, 151

Beyond these adjuncts, other potential techniques in treating fracture-related infection while maintaining the metalwork have also been published. These include the use of novel skin closure over irrigated negative pressure wound therapy dressing,¹⁵² and continuous local antibiotic perfusion (CLAP) to achieve and maintain high local antimicrobial concentration, exceeding the MIC to aid in FRI management.¹⁵³ Finally, there is an expanding body of research into the role of the gut microbiome on general health and specifically its impact on bone health.^154,155 The effect of antibiotic use on gut microbiology is well-known and accepted.¹⁵⁶ Considering the often prolonged use of antibiotics in FRI patients, monitoring and promoting gut health in a formal, regulated manner would seem to be a logical step in patient optimisation.

Although encouraging, identifying which therapies have true potential and translating that potential into the clinical environment is still a major challenge. Currently, many of these treatment methods lack robust clinical evidence and large-scale trials examining delivery, dosing, duration, and routes of application which are needed before they can be widely adopted. Realistically, there is no “silver bullet” (or silver-coated implant!), no one therapy that can provide complete management and resolution of FRI. Because of this, we foresee multimodal protocols using several adjuvant treatment approaches concomitantly as the future of FRI management, especially in stubbornly persistent infections and MDROs. Indeed, combining multiple different treatments simultaneously may be the key to combating and preventing the emergence of antimicrobial resistance in generations to come.

9. Conclusion

Trauma is inevitable, infection is inevitable and unfortunately, fracture-related infection is inevitable. What's more, fractures and infections affect every sector of society, irrespective of geographic or socioeconomic boundaries. As the absolute number of FRIs is certain to increase and as an estimated 10 million deaths from resistant bacteria is forecast in the next 25 years, it is imperative that the healthcare community organise, in particular trauma surgeons. PJI thought leaders have laid a blueprint for a progressive, systematic approach to localised infection and exceptional efforts have been made to standardise the management of FRI in a similar way. However, much work is still needed to promote sufficient funding and interest in high-impact journals comparable to PJI, along with staged innovation rooted in evidence and best practices. This review serves to inform the reader of the latest research, guidelines, and recommendations from FRI thought leaders on effectively diagnosing and managing fracture-related infections.

Funding

This research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Authors' contributions

EM Thompson: Manuscript preparation and writing.

AA Qureshi: Manuscript review.

Declaration of competing interest

None declared.