Osteomyelitis and Septic Arthritis of the Upper Extremity in Pediatric Patients

Osteomyelitis and septic arthritis are musculoskeletal infections with a variety of manifestations in children. Osteomyelitis is inflammation of bone secondary to microbial infection [1] and septic arthritis is microbial infection of the synovial membrane, joint space, and intraarticular structures [2]. The global burden of osteoarticular infections in children is immense, as annual hospitalization rates continue to increase around the world [3, 4]. Advances in the detection and treatment of both diseases have reduced the prevalence of long-term sequelae in many areas, but chronic disease and permanent disability continue to have devastating impacts on the health and quality of life for children in resource-limited areas across the world [5, 6].

Both osteomyelitis and septic arthritis tend to occur more frequently in the lower extremity (LE) compared to the upper extremity (UE) [7, 8]. Disability in LE infections is often more apparent given that weightbearing and mobility can be impaired. Perhaps due to smaller case numbers and an underappreciation of the disability caused by UE infections, less literature is available for those. As a result, the purpose of this review is to summarize the recent literature related to pediatric upper extremity osteomyelitis and septic arthritis and to provide up-to-date information regarding management of these conditions in children.

The overall incidences of pediatric osteomyelitis and septic arthritis have been reported to peak at rates of 80 per 100,000 and 10 per 100,000 respectively [9], but global trends vary significantly depending on geography, patient demographics, and microbial patterns.

The anatomical structure of the pediatric musculoskeletal system makes children more susceptible to disease compared to skeletally mature patients. Seeding of the bone and joint frequently occurs due to hematogenous spread but can also be the result of direct inoculation or contiguous spread [27‐30].

The most common sites for hematogenous osteomyelitis seeding are the metaphysis of rapidly growing bones. These areas are highly vascularized and contain leaky vessels with slow blood flow, which enables microbial deposition and proliferation in the capillary loops [31, 32]. In the UE specifically, the relative frequencies of osteomyelitis in the humerus, ulna, and radius are 5–13%, 1–2%, and 1–4% respectively [33]. Hematogenous septic arthritis occurs in a similar fashion, where microbial accumulation in highly vascularized joint synovium leads to infection [27]. Contiguously spread septic arthritis occurs most frequently in the shoulder, elbow, and hip because they have metaphysis that are intracapsular, which allows infections that breach the cortex of metaphyseal bone to enter directly into the joint space [30]. A recent multicenter study of 684 children in the US reported that 10% of septic arthritis cases occurred in the upper extremity, with 53% in the elbow, 41% in shoulder, and 4% in the wrist [17]. In a study that compared septic shoulder and elbow, children with septic shoulder tended to be younger, 1.0 year versus 4.6 years respectively [17].

The relationship between osteomyelitis and septic arthritis has been well described in the pediatric population. Rates of combined infections in the US increased from 0.8 to 1.3 per 100,000 between 1997 and 2012 [3], and a US multicenter study discovered that 46% of children with UE septic arthritis had adjacent musculoskeletal infections and/or bacteremia [17]. While a study of 247 European children from 2003 to 2018 reported that 17% of patients presented with concomitant osteomyelitis and septic arthritis, with 9.5% involving the humerus [25], these rates are lower than those reported in another European study, where concomitant bone or joint involvement was found in 24% of patients with osteomyelitis and 36% of patients with septic arthritis [34]. Higher incidences of concomitant infections are expected in very young children because they have transphyseal vessels that span the growth plate, which facilitate spread [29]. However, research highlighting that rates of concomitant infections among older children were only slightly less than those among younger children suggests that all age groups should be evaluated for concomitant infections [34].

Acute osteomyelitis cases often present within 1–3 weeks and are associated with inflammatory bone changes, while chronic infections last more than 4 weeks from the onset of symptoms and are characterized by the presence of necrotic tissue and sequestrum. Septic arthritis often presents earlier in onset due to significant pain in the joint from the onset of the infection [19, 27].

The goal of treating acute osteoarticular infections in children is to provide prompt therapeutic interventions that stabilize the patient, prevent disseminated disease, and avoid devastating long-term complications. This task is best undertaken with a multidisciplinary team composed of pediatricians, infectious disease specialists, musculoskeletal radiologists, pediatric orthopaedic surgeons, and ancillary staff [50]. Antibiotics serve as the foundation for the non-operative treatment of acute osteoarticular infections, and the most effective treatment regimens are based upon patient age, extent of disease, local sensitivities, and directed therapy with pathogen identification [46]. Historically, antibiotic therapy was initiated after culture samples were taken in an effort to improve cultural yields, but two retrospective studies questioned this, showing that tissue culture sensitivities of musculoskeletal infections were not significantly decreased by antibiotic administration prior to the collection of cultures [51, 52]. While this finding suggests that empiric therapy with broad spectrum coverage can be started immediately at the time of diagnosis, randomized-controlled trials are needed to confirm this approach.

The choice of empiric therapy continues to be an area of debate. For example, in separate studies of septic shoulder, the empiric antibiotics of choice were clindamycin in one study [18] and vancomycin in the other [32]. While both of these agents are effective against MRSA infections [53], an individualized approach that considers local guidelines, microbial resistance patterns, and patient factors should be employed when initiating empiric antibiotic therapy [31, 54]. Additionally, antibiotic therapy should be narrowed as soon as sensitivity data is available to limit unintended consequences of antimicrobial use [55]. Commonly used antibiotics in studies of UE infections include amoxicillin clavulanate [22], cefuroxime [38], and oxacillin–gentamycin [21]. A third-generation cephalosporin should also be considered alongside empiric antistaphylococcal therapy given the increased prevalence of K. Kingae in toddlers [15, 29].

In uncomplicated cases of osteoarticular infections, 3–4 days of intravenous antibiotics are recommended prior to transitioning to oral therapy, though this decision is contingent upon clinical improvement, reduction in inflammatory markers, and the ability for the appropriate antibiotic to be taken orally [31, 54]. 2–4 weeks of oral therapy is often appropriate if inflammatory markers normalize and clinical resolution is evident; otherwise, the course of antibiotics should be extended to 6 weeks total [31]. This aligns with studies of humeral osteomyelitis and septic elbow, where average durations of oral therapy were approximately 2.6 weeks and 4.4 weeks respectively [20, 22]. Alongside antibiotics, recent investigations have suggested the use of corticosteroids as an adjunctive therapy in the treatment of septic arthritis. A meta-analysis reported that corticosteroids shortened hospital stays, the total duration of antibiotic therapy, and days to CRP normalization [56], but a separate Cochrane review determined the evidence for corticosteroids in children with septic arthritis to be of low quality [57].

Operative treatment of osteoarticular infections includes drainage, decompression, irrigation, and debridement [30]. In addition to aiding in source control, enabling antibiotic penetration at the site of infection, and cleaning the surrounding tissue, surgery enables intraoperative collection of samples that can be used to identify caustic organisms and guide targeted antibiotic therapy [54]. This has been found to prevent disease progression; morbidity, such as osteonecrosis and cartilage damage; and mortality [16, 30, 54].

Septic arthritis has the potential to rapidly evolve into systemic disease, so urgent removal of inflammatory products is necessary [45], especially if pus is present [2]. This can be done with aspiration, arthroscopy, or arthrotomy [45]. Arthrotomy resulted in positive short-term outcomes and low complication rates in a study of UE septic arthritis [17], and this approach was also the most commonly used intervention in studies of septic shoulder [18, 32, 38], elbow [20, 21], and wrist [39]. However, arthroscopic lavage was shown to be equally effective in treating septic shoulder [38, 58]. Furthermore, a systematic review comparing arthrotomy and aspiration for septic shoulder and elbow determined that both approaches, followed by intravenous antibiotics, were capable of achieving good clinical results, though aspiration was over three times more likely to result in additional procedures [59].

For acute osteomyelitis, surgical drainage is often necessary to resolve infections that are not responsive to antibiotic therapy for 48–72 h or that develop complications (e.g. moderately large abscesses, concomitant septic arthritis, etc.) [16]. In a study of humeral osteomyelitis, surgery was indicated for 53% of the patient’s because they had radiological signs of subperiosteal collection or potential adjacent septic arthritis [22]. Surgical procedures depended on intraoperative findings but often consisted of subperiosteal abscess irrigation and drainage [22]. It is important to note that variations in surgical management of acute osteomyelitis are often institution driven, and that incorporation of various clinical factors in decision-making can affect rates of surgical intervention [60].

Additional procedures are warranted if the following situations occur subsequent to surgical intervention and antibiotic therapy: progression of clinical symptoms, continued drainage, or persistence of elevated CRP levels [16, 27, 30]. In a study of septic shoulder, where arthrotomy was exclusively used, MRSA-infected patients were significantly more likely to require additional operations to eliminate infection than those with non-MRSA infections [18]. Furthermore, a study of subperiosteal abscesses, including 11 cases in the UE, found that the combination of intramedullary decompression/debridement and abscess drainage decreased the risk for repeat surgical intervention when compared to abscess drainage alone [61].

Prompt diagnosis and treatment of osteoarticular infections are essential to achieving good outcomes and preventing complications [9, 29, 30]. Although treatment was initiated on the day of admission in 90% of children with humeral osteomyelitis, two patients developed multiorgan failure secondary to sepsis and a third developed a fixed flexion deformity secondary to avascular necrosis of the trochlea [22]. These three children, in addition to four others, had concomitant septic arthritis, which was associated with significant delay to presentation and a more severe form of disease [22]. In a study of UE septic arthritis, avascular necrosis, recurrence of infection, repeat irrigation and debridement, and a pathologic fracture all occurred in patients with concomitant osteomyelitis [17]. Another study of septic shoulder found high rates of concomitant disease, with 75% of patients having adjacent osteomyelitis on MRI [18]. They did not find an association between concomitant disease and poor outcomes, but they did discover that MRSA-infection, compared to other organisms, was a predictor of more severe disease course, negative outcomes, and adverse sequelae [18], which aligns with other literature [2, 6, 29]. High rates of concurrent osteomyelitis were observed in a study of septic elbow, where 58% of septic arthritis patients had osteomyelitis [21]. Delayed diagnosis in two of these patients resulted in readmission and repeat surgery to treat persistent symptoms [21], and a third patient became septic and died from multiorgan failure during their admission [21]. While no patients became septic in a separate study of septic elbow, 20% of patients were found to have elbow stiffness at their last follow-up [20].

In general, adverse outcomes are related to severe disease, delayed therapy (> 4–5 days from onset of symptoms), neonatal infections, and location of infection [9, 27]. Multiple studies of pediatric osteomyelitis have shown that complicated cases occur more frequently in the UE when compared to the LE [10, 62], which is likely due to children exhibiting a limp with LE infection and therefore receiving medical attention earlier in disease course [62]. Despite this, research into pediatric septic arthritis suggests that the hip has the worst outcome of all joints [2], especially in neonates [63]. In an effort to standardize care delivery and improve pediatric outcomes, classification systems, clinical care guidelines, and prediction tools that proactively identify osteoarticular complications and assess disease severity have been developed [62, 64‐67]. For example, one study showed that sepsis and hypergammaglobulinemia were associated with higher frequencies of late sequelae (e.g. reduced range of motion, permeant deformity, etc.) [68] while another discovered that failure to reduce CRP by 50% at hospital day 4 or 5 predicted both acute and chronic complications [69]. A prediction algorithm utilizing age, CRP, duration of symptoms, platelet count, and absolute neutrophil count (ANC) was capable of distinguishing septic arthritis with adjacent infection, such as osteomyelitis, from isolated septic arthritis [70]. The authors suggested that children meeting ≥ 3 of these criteria may benefit from preoperative MRI, which aligns with studies of septic shoulder and elbow that recommend preoperative MRI to detect concomitant osteomyelitis and assist in surgical decision making [18, 21].

Access to appropriate medical care and follow-up is often enough to prevent acute osteoarticular infections from progressing to chronic disease. While one study showed that a majority of children with acute osteomyelitis did not require follow-up beyond the initial treatment period [6], other studies recommend follow-up within 2–6 weeks of discharge to assess clinical improvement [27, 29], with prolonged monitoring reserved for patients with complications or infections near the growth plate [1, 30]. The prevention of long-term sequelae associated with chronic disease is paramount in avoiding permanent disability. Septic arthritis is often a surgical emergency given destruction of cartilage and its ability to rapidly evolve into systemic disease [21]. As a result of its acute disease course, evidence regarding the chronicity of this condition is limited when compared to osteomyelitis.

Persistent symptoms and the development of necrotic bone are hallmarks of chronic osteomyelitis [28]. The necrotic bone, referred to as the sequestrum, is surrounded by pus and reactive bone sclerosis. There is usually new bone forming as well, the involucrum. (See Fig. 4) Necrosis may appear 1–2 weeks prior to the development of these features [28, 33]. Rates of chronic osteomyelitis tend to be higher in resource-limited areas due to delays in access to care where acute osteomyelitis goes undertreated or without treatment [71]. An epidemiological study of osteomyelitis in Nigeria found 65.8% of the children had chronic disease, with 84% of those children having an initial onset of symptoms that were either neglected or mistreated [36]. In contrast, a study of chronic osteomyelitis in New Guinea determined that 73% of the cases were caused by inadequate treatment of open fractures [71]. 17% of the infections in this study were of the forearm, 10% were of the humerus, and 83% of the S. aureus strains were MRSA [71]. The authors of this study were able to achieve successful treatment in 95% of cases using radical debridement and antibiotics [71]. In a separate study in Rwanda, locally-made, biodegradable, calcium-sulfate bone cement pellets impregnated with antibiotics were successful in treating 95% of patients with chronic osteomyelitis, including 30 children [72], but there was not a comparison group of those treated without antibiotic impregnated calcium sulfate. Of note, no recent studies show improved outcomes in children with use of antibiotic beads/cement relative to other treatments. In cases of post-osteomyelitic forearm segmental defects (see Fig. 5), fibular strut with additional bone grafting has been shown to be an effective reconstructive method [73, 74], as it enables union and restores a majority of forearm function. Vascularized fibula flaps have also been shown to be a valid approach for treating UE post-osteomyelitic bone defects [75]. Bone transport and other segment defect filling techniques have also been employed, but there is not recent literature available to review. Further research into the factors contributing to the persistence of chronic infections in resource-limited areas is essential to developing prevention and treatment strategies that improve outcomes in patients and improve musculoskeletal outcomes for future generations.