Role of ultrasound in the diagnosis of paediatric acute osteomyelitis

Acute limb pain and swelling are a common entity in the paediatric setting, caused by various aetiologies, however in certain circumstances – such as acute osteomyelitis (AOM) – early diagnosis and management are essential to prevent serious sequelae⁽¹⁾.

Acute osteomyelitis refers to acute infection of the bone and/or bone marrow with a duration of infection prior to treatment of less than 2 weeks. In subacute osteomyelitis, the duration of infection prior to treatment is 2 weeks to approximately 3 months, while in chronic osteomyelitis the duration of infection before treatment is greater than 3 months. In paediatric AOM, the most commonly affected site is the metaphysis of the long bones, and patient management requires early diagnosis and prolonged intravenous antibiotic treatment for 6 weeks, either with or without surgical debridement⁽²⁾.

It is essential to suspect AOM clinically in a child with pain and swelling of the limb, followed by further evaluation with laboratory parameters and imaging in order to establish diagnosis in the early phase and avoid complications. In our study, we evaluated the role of ultrasonography (USG) in the early diagnosis of paediatric AOM, and also correlated the findings with magnetic resonance imaging (MRI) results.

The aims of our study were to evaluate cases of extremity pain or swelling in paediatric patients using USG to diagnose AOM, and to correlate USG findings with MRI findings.

A total of 18 paediatric patients referred to the radiology department of our hospital from July 2019 to July 2020 were evaluated. Following clinical assessment and laboratory work-up, the patients were sent to the radiology department for imaging evaluation. The clinical criteria considered to be indicative of AOM included pain and swelling of the extremity/joint, with fever, local tenderness, and/or redness. The laboratory criteria suggestive of possible AOM in a clinically suspected case of AOM were comprised elevated total leucocyte count (TLC), and elevated erythrocyte sedimentation rate (ESR) and C-reactive protein values (CRP). The initial radiographic examination of the affected limb was followed by USG and MRI scans.

A total 18 paediatric patients (15 boys and 3 girls) with a mean age of 8 years were evaluated

Out of 18 children, 7 children had no obvious clinical signs of infection, with essentially normal laboratory parameters, and radiographic examinations of the affected limb revealing alternative diagnoses, with fractures in 2 cases, osteosarcoma in 1 case, and osteoid osteoma in 1 case. In the remaining 2 cases in this group, USG findings showed a soft tissue collection which was away from the bone. Fine-needle aspiration cytology (FNAC) results in these cases were suggestive of soft tissue haematoma.

In the remaining 11 children, AOM was suspected on the basis of clinical findings including pain and swelling of the extremity/joint, with fever and local tenderness, and redness in a few cases. Elevated TLC, ESR and CRP values were present in most of these patients. Out of these 11 children, 6 had superficial collection and, based on FNAC results, 4 were diagnosed with cellulites, and 2 had soft tissue haematoma. Finally, a total of 5 patients (27.8%) were diagnosed with acute osteomyelitis meeting the USG criteria for AOM, which was also confirmed by MRI and surgical drainage. Therefore, out of 18 patients evaluated in the study, 13 patients (72.2%) had other causes of limb pain, as detailed above (Tab. 1).

The mean age of AOM patients was 8.2 years (the youngest child was 6 years, and the oldest 12 years old). Male children were more commonly affected compared to females (male 80%, female 20%). The femur was the most common bone involved (80%), and the most common site was the distal metaphysis. Tibial involvement was seen in only 1 case (20%). The right lower limb was more commonly affected, with 3 cases (60%), while the left lower limb was involved in 2 cases (40%). None of the cases of AOM in our study was characterised by upper limb involvement. The most common presentation was pain at the affected site (seen in all cases), while swelling was observed in 4 out of 5 cases of AOM (80%). No significant clinical swelling was seen in the case where the proximal tibia was involved. Redness and adjacent joint swelling due to synovial effusion were noted in 1 case out of 5 (20%). Fever was present in 2 cases (40%). The mean time of presentation to hospital and USG was 4.6 days (range: 3 to 7 days) (Tab. 2).

On the USG images, deep soft tissue fluid collection around the bone was present in all cases (100%). The collection showed internal echoes in all the patients. In a single case, significant internal echoes were seen, and the collection appeared hyperechoic. Periosteal thickening or elevation with subperiosteal fluid collection was seen in a total of 4 cases (80%). Increased vascularity within or around the periosteum on colour Doppler was noted in 4 cases (80%) (Tab. 3, Fig. 1, Fig. 2, Fig. 3).

Fig. 1.

Fig. 2.

Fig. 3.

All AOM patients underwent MRI examinations after the USG for the correlation of the findings, confirmation of diagnosis, and evaluation of extent. The MRI showed alteration of the normal marrow signal intensity in all 5 cases, and also subperiosteal collection in 4 cases, as seen in USG (Fig. 4). An additional finding of intramedullary collection was observed in 1 case of AOM.

Fig. 4.

Surgical drainage was performed in all patients (Fig. 5).

Fig. 5.

One of the most common causes of acute limb swelling, pain and joint effusion in children is AOM¹. It has been established over that years that on account of its peculiar vascularity the metaphysis of the long bones in children is particularly predisposed to AOM^(3–5). It is essential to diagnose AOM in the early phase by clinical examination complemented by diagnostic imaging (including conventional radiography, USG, computed tomography (CT), MRI and bone scan) in order to reach an accurate diagnosis^(2,6,7). The first-choice imaging modality in acute limb pain and swelling is plain radiography, being a readily available, relatively safe, inexpensive, and rapid technique. There are no definite changes seen in initial radiographs in AOM, and the role of radiographs is to establish alternative diagnoses rather than exclude AOM^(7,8).

CT is associated with radiation hazard, and thus it is not the modality of choice for marrow abnormalities as seen in AOM. However, CT visualises the cortical bone well, and thus has a role in the diagnosis of chronic osteomyelitis, foreign bodies or gas associated with infection^(9,10).

MRI is highly sensitive for bone marrow abnormalities and soft tissue infection, and highly accurate in the diagnosis of AOM. Our study also depicted changes of AOM in MRI in the acute phase. Additionally, MRI demonstrated the presence of intramedullary collection in 1 case, indicating that MRI is helpful in diagnosing complications of AOM more accurately than USG. Consequently, in view of more limited availability, high cost, and technical difficulties in the paediatric population MRI should be reserved as a problem-solving tool^(7,11,12).

USG is a rapid, cheap, easily available, non-ionising and fairly accurate diagnostic modality. It also helps to locate the lesion for diagnostic aspiration^(6,8). The first use of USG as a diagnostic imaging modality in AOM was reported in 1988 by Abiriet et al.⁽¹²⁾. The authors suggested that fluid around the bone seen on USG might indicate AOM. In 1994, Kang et al.⁽¹³⁾ demonstrated that the presence of subperiosteal collection/abscess as early as on the 4^th day of onset indicates the presence of AOM. In our study, all paediatric patients with limb pain or swelling were included, but only 27.8% were diagnosed with AOM. Previous studies^(12–15) demonstrated either deep soft tissue thickening or collection in most cases of AOM, while in our study the most important USG finding observed in all confirmed cases of AOM was deep soft tissue fluid collection around the bone. In our study, fluid was seen around the bone instead of deep soft tissue thickening, as the majority of cases were evaluated after 3 days of onset, whereas in most other studies the thickening on USG was seen during the first three days of onset. We also demonstrated periosteal thickening or elevation with subperiosteal fluid collection/abscess formation in most cases of AOM.

Over the years, with advances in ultrasound technology and improvements in resolution, the sensitivity and specificity of USG in diagnosing AOM has increased. The use of colour Doppler in the diagnosis of AOM was described by Chao et al. in 1999 in 12 paediatric cases of AOM⁽¹⁴⁾. In our study, increased vascularity within or around the periosteum was also seen on colour Doppler in most of the cases (80%). In 2005, Azam et al.15 published a study comprising of 55 children with AOM of limbs, where subperiosteal abscess was demonstrated in 42 cases (76.3%), while soft tissue swelling contiguous with bone was seen in all cases except for 3 patients in the early phase (1–3 days) of the disorder. The authors also demonstrated an increased vascular flow within or around the affected periosteum in all cases on colour Doppler. In their study, the most common site of involvement was the distal femur (45.4%), and our study is also consistent with this finding. Upper limb involvement was seen in 12 percent cases in their study, while none of our cases had upper limb involvement, which can be attributed to a smaller sample size in our study.

In 2011, Ezzat et al.⁽¹⁶⁾ published their study of 27 children with AOM, highlighting that AOM can be detected in the early phase by high-resolution USG and power Doppler sonography. They also showed the presence of cortical erosions in 48% cases of AOM, while none of our cases showed them, which is likely due to the fact that the scanning time in our study was between 3 to 7 days, and cortical erosions are normally seen after the 7^th day. In addition, the authors demonstrated the presence of synovial effusion in the adjacent joint in nearly 26% cases. In contrast, in our study only 1 case had an associated joint effusion.

Malicus et al.⁽¹⁷⁾ reported that the sensitivity of USG in diagnosing AOM was significantly lower than MRI in the early phase, however in our study we demonstrated that an early diagnosis of acute AOM was feasible with USG, especially after the 3^rd day of onset, revealing similar changes as those visualised by MRI. Therefore, being a cheap and readily available imaging modality, USG should be included in the treatment protocols of AOM to enable an early diagnosis.

We propose the following algorithm for the early diagnosis of paediatric AOM by USG (Fig. 6):

Fig. 6.

AOM is a common entity in the paediatric population, presenting with acute limb pain and swelling. Early diagnosis and management of AOM are essential to prevent serious complications. USG can play an important role in the early diagnosis of paediatric AOM, and should be incorporated into treatment protocols followed in cases of suspected AOM. MRI should be reserved as a problem-solving tool.