154 SECTION Ill Common Orthopaesic Disorders described nonspecific inflammation, with a predominance of lymphocytes, plasma cells, histiocytes, and multinucle- ated giant cells. Evaluation The diagnosis of CRMO is essentially one of exclusion. The following criteria have been suggested to avoid the pitfalls ‘of misdiagnosis: (1) protracted clinical course of greater than 3 months’ duration, (2) open biopsy results consistent with chronic inflammation, and (3) failure to identify any infectious organism by culture," ‘The initial evaluation of CRMO should include the same laboratory and radiographic studies performed for any suspected infection. Microbiologie tests of blood and tissue cultures should not be omitted merely because CRMO is suspected. Plain radiographs of each symptom: atic area should be obtained and supplemented by bone scintigraphy because asymptomatic lesions are occasionally identified in this manner'"'"" Areas of lysis and reactive periosteal sclerosis can mimic conditions such as Ewing sarcoma, metastatic neuroblastoma, and Langerhans cell histioeytosis."” Ifa malignant neoplasm is considered in the differential diagnosis, biopsy principles should be followed carefully when obtaining a specimen for histopathologic Treatment CRMO has no well-defined treatment, although most chil dren are treated with NSAIDs and occasionaly corticoste roids. Patients should not respond to antibioti treatment. fective treatment with interferon-y was reported in one ‘ase, and itis hoped thet this method may help circumvent the prolonged, relapsing nature of CRMO. A 13-year-old srl who was treated for 3 months with interferon-y had no symptomatic episodes in the 15 months immediately after treatment, compared with 11 symptomatic episode: 24 years before treatment." The exact mechanism for beneficial effect is not known, but the use of interferon-y in cases of chronic granulomatous disease (CGD) and in the treatment of intracellular pathogens such as in leprosy, M avium-intracellulare infections, and leishmaniasis, which are often refractory to standard antimicrobial treatments, may offer indirect support to the theory of an obscure pathogen as the cause of CRMO. Most children show 2 favorable clinical response to NSAIDs.” Bisphosphonates, particularly pamidronate have been reported to have a dramatic favorable impact on pain, vertebral remodeling, and. skeletal inflammation on MRI in refractory cases.°""""" Severe and_ prolonged disease can occur in a minority of patients despite intensive treatments.” Complications CRMO is often described as a self-limiting condition of childhood that follows a benign, protracted course without sequelae.“ An early diagnosis can avoid complications associated with unnecessary surgical or antibiotic treatment. However, two long-term follow-up studies suggested that the natural history of CRMO is prolonged and associated with adverse sequelae (deformity, limb length inequality and pain) m some individuals." Both reports suggested ‘that overall, most individuals do well both physically and emotionally despite the long-lasting effects of the illness." Septic Arthritis Epidemiology Septic arthritis accounts for approximately 0.25% of hospi- izations among children. The condition is thought to be more common than osteomyelitis. One retrospective review identified 471 cases of septic arthritis over 26 years, com- pared with only 258 cases of osteomyelitis in a correspond: ing 22-year period." The same review noted that most infections (70%) were identified in children from 1 month to 5 years of age, ani half of these infections occurred in children younger than 2 years. Single joint infection was noted in 94% of children, most commonly involving the hip (419%) and knee (23%), followed by the ankle (14°), elbow (12%), wrist (4%), and shoulder (4%). Septic sacroilitis, although rare, has been reported in children and is often overlooked because of the vague clini presentation." The spectrum of causative bacteria and the frequency of occurrence of specific pathogens are similar to those seen in osteomyelitis, with S. aureus being vhe most common organism identified. Certain bacteria appear to have a higher likelisood of causing septic arthritis than of causing osteomyelitis. These organisms include B. melitensis, H. influenzae, K. kingae, N. meningitidis, and NN. gonorrhoeae." ?""*** Septic sacroilits is associated with Mycobacterium tuberculosis." Pathophysiology Hematogenous seeding of the synovium during transient bacteremia is the most common cause of septic arthritis in children, Other foci of infection, such as otitis media or sinusitis, may be present well in advance of the joint infection." Septic arthritis may also arise from a contigu- ous site of infection, such as adjacent osteomyelitis, oF from a penetrating injury with direct inoculation into the joint.” Joints that are particularly susceptible to the spread of infection from an adjacent osseous source include the knee (31%), hip (23%), ankle (18%), and shoulder (14%); this susceptibility is largely the result of the intracapsular location of the metaphysis in these joints (Fig. 27-23)" Bacterial entry into a joint space signals the onset of an inflammatory cascade that, left untreated, may lead to car silage destruction and loss of normal joint function. The precise pathway of joint degradation is not fully understood. However, research suggests that macrophages, polymorpho- nuclear leukocytes, and synovial cells release cytokines (IL-1B, IL-6, TNF-a), immunoglobulin G, and lysosomal + into the joint space.’”" This process results in an 388 of proteoglycan subunits from the cartilage matrix, which may be severe as early as 2 to 5 days after the onset of infection despite the lack of visible cartilage degenera- tion. In an experimental model of septic arthritis in rabbit knees, S. aureus injection resulted in proteoglycan subunit loss of 30% at 48 hours, SO% at 5 days, and 80% at 3 weeks; collagen degradation did not ensue until 3 weeks (28% loss). It remains unclear whether joint cartilage is able toc FIGURE 27-23 Metaphyses of the proximal radus (A) proximal humerus (B), proximal femur (C), and distal tibia and fibula (D) are inraarlicolar. Osteomyelitis in these localions may decompress into the joint and produce concomitant septic achrite. restore normal proteoglycan content after elimination of the bacterial infection and before the onset of collagen loss Some strains of S. aureus possess a gene encoding for collagen-binding adhesion (Cra). Experimental injection of aureus in mice resulted in septic arthritis in 70% of animals injected with Cna-positive strains, versus only 27% of animals when the gene was absent." Evaluation Septic arthritis should be considered whenever an ill appearing child with a clinical history of atraumatic limita- Lion of mobility has the physical finding of joint irritability. use the same initial diagnostic process for septic arthritis as for osteomyelitis: plain radiographs and laboratory studies (CBC with differential, CRP ESR, and blood cultures) When symptoms are located in the hip region, comparative ultrasonography is performed to evaluate for hip effusion. The differential diagnosis of septic arthritis includes tran. sient synovitis, reactive arthritis, juvenile rheumatoid arthritis, Kawasaki syndrome, Henoch-Schonlein purpura heumatic fever, avascular necrosis, slipped capital Femoral epiphysis, trauma, neoplasia, Lyme arthritis, and Legg: Calvé-Perthes syndrome.” Other infections occurring near a joint, such as osteomyelitis, pyomyositis, septic bursitis, cellulitis, and abscess, can mimic the clinical presentation of septic arthritis. In most cases, st should be possible to narrow the differential diagnosis of an acutely irritable joint significantly following the inital evaluation. Despite all the information that can be obtained with modern laboratory and radiographic studies, itis important not to lose sight of | she value of obtaining a complete history and performing @ shorough physical examination when dealing with musculo- skeletal infections. Septic arthritis can be differentiated from osteomyelitis by the presence of a warm joint that is "Refezences 7, 8 24, M1, 131, 265, 285,283, 345,405, 436. (CHAPTER 27 Infections ofthe Musculoskeletal System 1055, painful with gentle passive motion; joint motion usually does not exacerbate symptoms in osteomyelitis.” Frequently, the greatest diagnostic challenge is differen- tiating between septic arthritis and transient synovitis of the bip. Because of the importance of correctly identifying these two conditions, great attention has been focused fon devising an evidence-based clinical prediction strat egy Tailure to identify septic arthritis correctly ‘may result in poor long-term outcomes." "Transient synovitis is one of the most common causes of hip pain in children; itis responsible for up to 0.9% of pediatric emer- {gency room visits each year, and early diagnosis can avoid unnecessary invasive procedures and hospitalization for observation.” Kocher and colleagues identified four independent pre dictors to help differentiate between septic arthritis and transient synovitis: history of fever (oral_ temperature >38.5°C), history of non-weight bearing, ESR greater than 40 mm/he, and WEC count greater than 12,000 cells/ml. Although individual variables alone were not useful in dif ferentiating between these conditions, the authors were able to demonstrate through multiple logistic regression analysis that the predictive probability of septic arthritis in the population studied was 0.2% for 0 predictors, 3.0% for 1 predictor, 40% for 2 predictors, 93.1% for 3 predictors, and 99.6% for 4 predictors.” This method was prospec tively validated at the same institution by studying 213, children between 1987 and 2002," The authors found that the actual distribution of septic arthritis wae 2% for 0 pre dictors, 9.5% for 1 predictor, 35% for 2 predictors, 72.8% for 3 predictors, and 93% for 4 predictors This same clinical prediction algorithm was used in a retrospective review by Lukmann and co-workers but dem ‘onstrated only a 59% predicted probability of septic arth tis when all four independent variables were identified." ‘The authors attempted to select a better model by choosing thrce alternative predictors: history of fever, serum WBC ‘count greater than 12,000/mm', and a previous health care visit; however, the predicted probability of septic arthritis vwas only 71%. On the basis of this low predictive probabil ity, these authors recommended the use of hip ultrasonog raphy and arthrocentesis as adjunctive diagnostic modalities in the evaluation of an irritable hip. Building on the work ‘of these previous authors, others have found that fever (oral temperature >38.5°C) was the best predictor of septic arthritis." ‘The valuable work from these two institutions demon- strates the limitations of relying excessively on clinical predictors; these predictors inevitably have diminished performance in a new patient population beeause they were established to model the original population studied," °**"" However, the presence ot absence of multiple independent predictors of septic arthritis in any given patient can be helpful in making the decision whether to observe the patient or aspirate the hip joint in the operating room with the patient under general anesthesia, Ultimately, no subst tute exists for vigilant surveillance and good clinical judgment. ‘Many authors consider aspiration of a septic joint to be a significant and indispensable part of the diagnostic process." "28" Aspiration of the hip joint may be difficult, but a variety of methods can be used to access thisA aspiration attempt joint safely for arthrocentesis, including ultrasound and fluoroscopic guidance." Although blind aspiration based on. anatomic landmarks can be attempted, i is not recom ‘mended because itis impossible to confirm an intraarticular location in the ease of a negative aspirate, as can be done ‘with ultrasonography or an arthrogram (Fig. 27-24). Joi fluid obtained by aspiration should be sent for Gram stain and culture, as well as cell count. Some specialists also ‘obtain joint fuid glucose and protein levels and compare these with serum levels. A WBC count greater than 50,000/ mm! with a predominance of polymorphonuclear leuko: cytes, a high protein content, and a low glucose concentra tion (<33% of serum glucose) are characteristic of sept arthritis. The information obtained from the Suid cell count is important in the diagnostic process because joint fluid ‘may inhibit the growth of certain bacteria and prevent the positive identification of an organism. Although most authors report a low rate of culture-negative septic arthritis (ranging from 18% to 48%), one study reported a rate ae fighas 70457200 Treatment Antibiotic Therapy Empiic antibiotic selection for septic arthvtis is similar to that for osteomyelitis. For adolescents suspected of having disseminated gonococea illness, ceftriaxone should be con sidered initially. Peak synovial fluid concentrations of com- monly used antibiotics have been shown to be greater than 650% of peak serum concentrations, with adequate inhibi- tory activity against the common organisms that cause septic arthritis.” Sequential parenteral-to-oral antibiotic therapy is well established as an effective method of treat rent and reduces hospital stay, cost, morbidity, and incon- venience to families." Evidence suggests tha chiléren ‘who were treated with an early transition to oral antibiotic hhad a clinical response and an outcome equivalent to those in children who had late transition to oral therapy.-”*"” In {eneral, he duration of antibiotic therapy for most cases of uncomplicated septic arthritis is approximately 4 weeks, FIGURE 27-24 Fluoroseopieally guided aspiration ofa hip jint (A) with an arthrogram (B) to conti the intraarticular location of the — ‘Common Orthopaedic Disorders & although shorter treatment courses have been described. As in the treatment of osteomyelitis, I decide whether to extend the duration of treatment based on normalization of, the ESR. Surgery Tt is commonly agreed that some form of joint decompres- sion with aspiration and lavage, repeated aspirations, arthroscopy, or open arthrotomy should be performed as the inital treatment for septic arthritis, along with the intravenous administration of an appropriate antibiotic." Debate continues over which of these methods is best and the appropriate time frame for the intial invasive interven- tion. The most conservative approach involves performing joint aspiration in the emergency department with the use of conscious sedation or in the operating room with the patient under general anesthesia, The expectation is that joint arthrotomy will be performed immediately as an emergency procedure ifbacteria are present on Gram stain, the joint fluid cell count is greater than $0,000/mm', or the clinical suspicion of septic arthritis remains high, regardless of the joint fluid findings, Other methods of treatment have also proved to be successful. Serial joint aspiration, which 4s well documented for knee and shoulder joint sepsis, was studied in the hip joint, and surgery was avoided in 24 of 28 patients ina reported series.” The mean number of aspirations in that study was 3.6 per child, and 759 of children resumed walking after 24 hours ‘A growing body of evidence attests to the benelits of arthroscopy in the treatment of septic arthritis of the hip, knee, ankle, shoulder, and elbow. Several advantages of this method have been reported, including the following comprehensive visualization of intraarticular structures improved ability to assess the severity of sepsis affecting the joint cartilage and synovium compared with limited arthrot- ‘omy; capability to remove Rbrinous aggregates or advanced synovitis, which may serve as a source of persistent ‘References 35,75, 208, 240, 258,262, 287 "References 81, 84, 287,371, 379, 384, 417(CHAPTER 27 Infections ofthe Musculoskeletal System 1057 FIGURE 27-25 Arthroscopic image of a knee jaint 3 days after arthrotomy, irrigation, and drainage of septic arthits. The orginal antrotomy was through a small Incision, with limited visualization ofthe joint. Lack ofa clnial and laboratory response aftr the intial {urgery led to tie fllov-up procedure. Frdings at arthroscopy included 3 cifusefibrinous cast of the knee and synovitis invahing the Suprapatellar pouch and medial and lateral femorotial gutters (A), Aker arthroscopic debridement (B and ©), the child demonstrated rapid clinical and laboratory improvement. infection if left unattended; allowance of early functional rehabilitation because of the minimally invasive approach; and ease of accurate drain placement through arthroscopic cannulas to ensure thorough postoperative drainage (Fig 27S. Tg seties of 76 children with scopicaly ented septic ats, 91% were cured by arthroscopic irigation and antibiotis alone, and ops sion was required in only 4%." “Although the necessity of emergency intervention has rot been clearly established on the basis of clinical oF hi topathologic evidence, several studies suggested that ear and aggressive intervention is beneficial.” "* Satisfae tory clinical results were reported as long as intervention occurred within 4 days of the onset of symptoms.” Treatment delay has been associated with fibrinous locula- tions, synovitis, and pannus formation, which may lead to persistence of infection, further surgery, and prolonged hospitalization.” Clinical Practice Guideline Based on a systematic review ofthe best available evidence, an interdiseiplinary committee at Childeen’s Hospital in Boston developed a clinical practice guideline for the treat- ment of septic arthritis ofthe hip in children." Comparing a historical control group of 30 children with septic arthritis, with a similar prospective cohort group, the authors found no significant difference in clinical outcome.”* However, they noted a significant improvement in the standardization of care for the children treated under the guideline, which resulted in greater compliance with recommended an otie therapy (93% versus 7%), faster change to oral antibiot= ics (8.8 versus 69 days), and shorter hospital stay (4.8, versis 83 days). Although outcome variables remained unaffected in this limited series, one can only assume that a positive effect would be identified in a larger series, particularly if the guideline could prevent unnecessary delays in treatment or inadequate treatment, which could otherwise occur without evidence-based practice standardization, Complications Complications associated with septic arthritis include systemic sepsis, premature arthrilis, osteonecrosis of the proximal femur (Fig. 27-26), physeal closure, growth dis: turbance, synovitis, arthrofibrosis, joint stiffness, and per sistent infection, A delay in treatment has been found to be the single most important factor affecting the prognosis in children; no unsatisfactory results occur when symptoms are present for less than 3 days before treatment. prognostic factors include anatomic location of the infec tion, presence of adjacent osteomyelitis, and adequacy of, treatment.” It is well recognized that the hip joint is vulner able to complications, particularly when infection occurs at an early age, when the anatomy allows rapid communication of infection from the metaphysis to the epiphysis and into the joint (see Fig. 27-26)."* Poor results are associated with, delay in definitive treatment longer than 5 days and the presence of osteomyelitis of the proximal femur (Four of Tuberculous Arthritis ‘Tuberculosis of bones and joints is a granulomatous inflam. mation caused by M. tuberculosis. It is a localized and destructive disease that is usually blood-borne from a primary focus such as infected peribronchial or mesenteric lymph nodes, typically involving metaphyseal spread of M. tuberculosis into the joint.” This transphyseal spread is characteristic of tuberculosis and is not seen in patients with, pyogenic arthritis, This route of infection is particu prevalent among children younger than 18 months of age, at which time the transphyseal vessels disappear and exten: sion of infection into the epiphysis and joint becomes less ‘common. The infection may be of the human or the bovine type. In counteies where raw milk is used extensively, ovine transmission is common, whereas in areas where ilk is pasteurized and there is rigid control of dairy herds, the bovine type is extremely rare and the human type i¢ ‘After a period of marked decline of tuberculosis, espe. cially in North America, a gradual increase in the incidence ofthe disease hasbeennotedsincethelate 1980s." This increase has been closely associated with the acquired immunodeficiency syndrome (AIDS) epidemic. "References 21,36, 75, 78, 121, 164,195, 271, 40,