Due to the historically low resolution of imaging technologies, descriptions of septic arthritis of the facet joint (SAFJ) in children are scarce, although severe cases are known.
SAFJ was first mentioned in a French national study conducted by David-Chaussé et al1 in 1981, as a case among 2166 adults with bone and joint infections (BJI).
The first case report of a 66-year-old man with devastating late-developing bone lesions shown by computed tomography (CT) was published in 1987.2 The first pediatric case report was published much later, in 1995: a 10-year-old boy age. The child was diagnosed with JFAS by magnetic resonance imaging (MRI), which also revealed an epidural abscess3.
The characteristics of SAFJ are still poorly described and there is currently no consensus on its management; This lack of knowledge is a major barrier to diagnosis. To shed light on SAFJ in children, we first aimed to estimate the incidence of SAFJ in the pediatric population of the city of Lyon (France).
Then, by combining cases from the hospital database with those from the literature, the authors aimed to specify clinical, imaging, and laboratory findings, as well as identify pathways for appropriate management.
| Methods |
> Systematic review of published cases and guidelines
Systematic reviews were conducted, according to PRISMA guidelines, 4 to identify published cases of pediatric AFS and identify guidelines for AFS and BJI; among the latter, the criteria of the AGREE guidelines were used to select more relevant guidelines.5
> Unpublished cases from a 10-year consecutive case series
A systematic case review (2008-2018) was carried out according to the CARE6 guidelines at the Hôpital Femme-Mère Enfant, the only tertiary-level pediatric teaching hospital providing medical care to the 301,246 children of the city of Lyon. 7
> Data collection and analysis
The incidence rate of JFS in children was estimated for the general population of the city of Lyon (1,370,678 inhabitants), 7 expressed as a number of new patients diagnosed with JFS/100,000 children/year. The same clinical and care data were extracted from the articles (published cases) and hospital medical records (unpublished cases). Variables are expressed as medians (ranges) and percentages. The appropriateness of treatment based on selected pediatric guidelines for BJI was discussed.
Variables are expressed as medians (ranges) and percentages. The appropriateness of treatment was discussed based on selected pediatric guidelines for BJI.
> Ethics
The study protocol was approved by the ethics committee of Hospices Civils de Lyon (October 24, 2018). In accordance with current French regulations, parents were informed of the study. The database was registered with the French data protection agency according to protocol MR003 (CNIL-18-228, April 9, 2018).
| Results |
> Systematic review of published cases and guidelines
The systematic review recovered 10 articles that reported one pediatric SAFJ each, 3,8–16 and 4 selected guidelines17-20; no SAFJ guide was recovered.
> Unpublished cases
Among the 2867 spinal MRIs performed over the past 10 years, 105 were performed in an infectious setting; These cases underwent rereading to ensure there were no misclassifications, and a total of 7 children (ages 20 months to 17 years) were diagnosed with FAJS over 10 years.
The mean ± 0.4 annual incidence rate of JFS in children was 0.23 ± 0.4/100,000 children/year. During the study period, the number of cases identified increased over time, with 6 cases diagnosed in the last 3 years. Additionally, the number of spinal MRIs performed per year increased 2.6-fold. As case 7 did not consent to participate, the characteristics of 6 cases are reported.
> Case analysis
• Clinical presentation
A total of 16 cases are presented: 11 boys and 5 girls, between 18 months and 17 years of age [median (range): 7.5 years (1.5-17)].
Most cases were febrile (14/16, temperature between 37 and 39.5°C). They complained of acute low back pain (16/12) and/or lameness (16/6). Of the 6 unpublished cases, 5/6 children avoided or rejected the sitting position due to pain. Interestingly, the parents of case 3 reported the child’s refusal to sit on the potty. This particular information, available for 9 patients, was positive in 7/9 cases (78%).
The time from the appearance of the first symptom to diagnosis ranged from 1 to 28 days (median: 4.5 days). Local examination found lateralized pain (14/16), paravertebral inflammatory swelling (4/16), and/or spinal rigidity (12/16).
Neurological abnormalities were reported in 4 cases: right foot held in plantar flexion (1/16), right knee extension weakness (1/16), and sciatic pain (2/16). No septic shock occurred (0/16).
• Imaging: diagnosis and complications
All but one case underwent spinal MRI (15/16). In the 6 unpublished cases, MRI highlighted widening of the facet joint space with effusion, abnormal signal from capsule and ligament structures with hypointensity on T1-weighted images, hyperintensity on T2-weighted images or short IT inversion-recovery sessions, and enhancement after gadolinium injection, centered on the spinal facet joint.
Spine involvement was lumbar (14/16) and 2/16 were thoracic. The main differential diagnoses were spondylodiscitis and vertebral osteomyelitis; They were discarded due to lack of disc space and involvement of the vertebral body.
All SAFJs extended into the paraspinal muscles (15/15) and were complicated by pyomyositis in 2/16 cases. The SAFJ extended forward into the epidural space in 8/16 cases, including 5 epidural abscess complications (5/16).
MRI was repeated after completing antibiotic treatment in 6/15 cases. A decrease in paraspinal muscle hyperintensity was evident on short IT inversion-recovery images and inflammatory joint changes persisted around the facet joint at 3 days for case 5, 2 weeks for case 1, and 4 weeks for cases 16 and 3. In cases 13 and 2, the MRI normalized 4 weeks after ending antibiotic treatment.
Standard radiographs were performed in the majority of cases (13/16) and were considered normal. Ultrasounds were performed through the posterior spine and paraspinal area in 2/16 cases, finding a hyperechogenic lumbar paraspinal muscle infiltration (case 4) or an abscess of the paraspinal muscles (case 5).
CT was performed in 6 cases; In case 5, 21 days after the onset of symptoms, CT found bone erosion in the right facet joint T12-L1 and guided a local percutaneous aspiration biopsy. Scintigraphy was performed in 5/16 cases; In case 6, a nonspecific increase in uptake was found at the L5-S level.
• Laboratory data, bacterial identification and etiology
In the majority of cases (12/14), there was neutrophilic leukocytosis. C-reactive protein levels were often abnormal (>5 mg/L, 12/14).
In 4/16 cases, there was bacterial identification through positive blood cultures due to growth of Streptococcus pneumoniae (case 9), Streptococcus pyogenes (case 1), Staphylococcus aureus (case 2) and Staphylococcus epidermidis (in the absence of immunodeficiency, previous surgical or traumatic event, The role of a coagulase-negative Staphylococcus strain in causing the infection is debatable, and it is generally considered as a contaminant of the skin flora).
However, the article reports 2 positive blood cultures, supporting this etiology (case 10). In case 3 Enterococcus faecalis was not considered as the etiology of AFS (given the lack of urinary leukocytes and the recently positive urine culture).
In 3/16 cases, a local percutaneous aspiration biopsy was performed under CT guidance, identifying S. aureus (cases 13 and 5) and K. kingae (case 15) by reverse transcription-polymerase chain reaction (RT-). PCR). In 9/16 cases, no bacteria were identified.
Positive blood cultures confirmed hematogenous acquisition in 4/16 cases. Cases 16, 2, 4, 5 (4/16) reported a history of indirect trauma and case 13 (1/16) a direct trauma . Case 5, who presented with an epidural abscess, had a history of non-steroidal anti-inflammatory drug treatment for back pain during the 5 days prior to diagnosis.
> Treatment
All 16 patients received antibiotics for a median (range) duration of 37.5 (25-49) days. The mean duration of intravenous and oral antibiotic therapies was 9 (4-42) and 21 (14-42) days, respectively. Compliance with pediatric guidelines was acceptable for empirical antibiotic therapy (79%) and insufficient (62%) for targeted antibiotic therapy.
All treatments were followed by complete recovery (15/15) without sequelae (follow-up ranged from 29 to 69 months for unpublished cases). Among the 6 unpublished cases, apyrexia was achieved in less than 48 hours in all cases (6/6) and the children were pain-free in a mean of 5 days.
The unpublished cases were treated according to the French pediatric guidelines, 17 and validated during the weekly collegiate meetings on infectious diseases, except for case 5. Surgery was necessary twice (drainage of the pyomyositis abscess for case 15 and epidural drainage of the abscess for case 12).
| Discussion |
> Clinical presentation
In the setting of acute febrile low back pain and/or febrile lameness, suggesting acute BJI, any pain in sitting position, or "potty refusal," the non-verbal equivalent in young children, leads to consideration of AFS, spondylodiscitis21, or sacroiliitis. .22
When lateralized spinal tenderness and/or paravertebral inflammatory edema are associated with pain on sitting, the clinical picture supports an orientation away from spondylodiscitis and sacroiliitis and is sufficient to suspect JFS at the bedside and before any type of image
. As acute complications consist of an extension of the infection with or without abscess into the epidural space, observed in half of the cases, and into the paraspinal muscles, observed in all cases, the clinical examination should carefully detect any central neurological signs. (i.e. pyramidal syndrome in case of spinal cord compression) or peripheral (i.e. radicular or cauda equina syndrome), and these must be distinguished from low back pain and lameness. The detection of radicular pain is a sign of epidural involvement, previously described by Heusner23 as the second stage of the natural course of epidural abscesses, preceded by weakness (stage 3) and paralysis (stage 4).
> Images, diagnosis and complications
MRI appears to be the best imaging technique for diagnosing AFSJ; allows a positive diagnosis to be made 48 hours after the onset of symptoms 24 and to rule out differential diagnoses such as spondylodiscitis. All cases showed inflammatory changes in the paravertebral muscles at diagnosis (corresponding to clinical paravertebral edema) that have been suggested to be concomitant with facet joint infection due to common perfusion.25
In the setting of inflammatory changes in paravertebral soft tissues, radiologists must meticulously evaluate bony involvement to avoid misdiagnosis of AFSJ as primary myositis. MRI also allows the detection of complications such as involvement of the epidural space or a muscle abscess.26
Since inflammatory changes have been shown to persist for several weeks despite clinical recovery,27 in the absence of epidural abscess, MRI monitoring is not necessary until 6 weeks after antibiotic administration, unless new neurological signs occur. provide an urgent indication for monitoring.
Of note, the parallel increase in diagnosed FJAS over time and the increase in spine MRIs performed suggests a possible link between MRI availability and the ability to diagnose FJAS.
During the period 2007-2017, the number of MRI units/million inhabitants increased by 50% in Germany, Canada and the United States, 130% in Israel, 160% in France, 170% in Australia and 190% in Poland.28 Therefore, it is predictable that FAJS will be diagnosed with increasing frequency, which requires wide dissemination of knowledge about FAJS.
In settings where MRI is not readily available, simple clinical signs suggestive of AFSJ should provide adequate justification to refer a child to a larger center or to shorten the delay in obtaining MRI.
Relative to other imaging strategies, CT has the ability to diagnose only advanced bone or joint destruction and contiguous soft tissue abscesses and, therefore, cannot be considered a first-line image, but has great utility in guiding local percutaneous aspiration. Indications for scintigraphy should be limited to situations and settings where MRI is not available within an acceptable time frame.
> Laboratory data, bacterial identification and epidemiology
The bacteria found in the present study were the most frequent in bacteremia in children, corresponding to those expected in pediatric BJI.29 The proportion of positive blood cultures found was quite low , less than a quarter; this is likely due to the delay between bacteremia and the onset of BJI symptoms that trigger medical treatment, including blood cultures. S. aureus is the most common bacteria in children after the age of 5 years, as well as in adults.30 K. kingae is the bacteria most frequently involved in pediatric BJI between 6 months and 4 years.31
Since blood cultures often remain negative for this bacteria, they must be specifically targeted by RT-PCR for identification, preferably using surgical or aspiration samples.32,33 This is particularly well illustrated in the series reported by Ferroni et al29; in which, despite an even lower positivity rate of blood cultures (10%), joint aspiration (joint fluid was removed from patients with joint infection) was systematically performed (90%) and a final rate of 63 was obtained % identification (1.5 times higher compared to the present study).
> Treatment
In adults, the choice of antibiotic type has been discussed by Ergan et al. 25 and Muffoletto et al.,30 suggesting that the pattern designed for spondylodiscitis is also effective in AFSJ.
However, unlike adults, 19,20, there are no pediatric guidelines for the treatment of spondylodiscitis and treatments should follow general guidelines for BJI. According to the European Society for Pediatric Infectious Diseases, short intravenous therapy followed by oral therapy is appropriate in most children with uncomplicated BJI.18
The French pediatric infectious diseases group of the national scientific society (Groupe de Pathologie Infectieuse Pédiatrique — Société Française de Pédiatrie) recommends empiric antibiotic treatment for spondylodiscitis with adequate coverage against (1) methicillin-susceptible S. aureus, (2 ) streptococci (S. pyogenes and pneumococcus) and (3) K. kingae with high-dose intravenous ampicillin/clavulanic acid (150 mg/kg/day) or cefamandole,17 although the European Society for Pediatric Infectious Diseases guidelines promote first or second generation cephalosporin.18
Since current data confirm that the bacteria involved in SAFJ do not differ from those found in other BJIs in children, first-line therapy should be the same as that recommended for other BJIs. The initiation of treatment (empirical antibiotic therapy) was fairly consistent with the guidelines, but the tailoring of treatments could have been improved by narrowing the spectrum and switching to oral antibiotics.
Regarding the duration of antibiotic therapy, although pediatric guidelines do not specify its duration for children, adult guidelines recommend a 6-week treatment for spondylodiscitis.
In unpublished cases, a median of four to five weeks of antibiotic therapy allowed complete recovery without failure or relapse, suggesting that pediatric JFAS may require a shorter duration of antibiotic therapy than adults with spondylodiscitis; The improved diffusion of antibiotics, related to the improved vascularization of the facet joint compared to the amorphous intervertebral disc, could support this observation. Shorter courses of antibiotics have been recommended for several infections, as this can potentially reduce antibiotic resistance.34,35
| Conclusions |
To date, the present work constitutes the largest database of SAFJ in children. The mean incidence rate of JFS in children was estimated to be 0.23 ± 0.4/100,000 child-years.
The key clinical symptoms were refusal to go to the toilet in young children or pain when sitting, combined with lateralized signs. When these 2 signs are present, early suspicion of AFSJ is possible at the patient’s bedside.
Confirmation of the diagnosis and evaluation of extension to the epidural space is achieved by MRI. Treatment of SAFJ is based on targeted antibiotic therapy based on blood cultures and PCR for K. kingae from joint aspirates for children younger than 5 years.
As the availability of spinal MRI increases, SAFJ detection will likely skyrocket. A broader understanding of the characteristics of pediatric JFS is necessary to provide early diagnosis and optimize disease management.
