Clinical Characteristics and Management of Children and Adolescents Hospitalized With Pyomyositis

Background: Pyomyositis, a bacterial muscle infection, is an important differential diagnosis in children and adolescents with musculoskeletal pain. In contrast to tropical regions, it is rarely recognized in temperate countries, but incidence is increasing and major studies are missing. Methods: This retrospective multicenter study included patients <18 years of age hospitalized with pyomyositis in 11 Swiss children’s hospitals between January 2010 and December 2022. Cases were identified by ICD-10 code (Myositis; M60–M60.9), and data was extracted from electronic hospital records. Results: Of 331 patients identified, 102 fulfilled the case definition. Patient age at presentation ranged from 2 weeks to 17 years (median 8 years). The majority had no underlying illness and all presented with fever and localized pain. At the respective site of pyomyositis, 100 (98%) had impaired movement and 39 (38%) presented with local swelling. Pelvic (57%) and leg (28%) muscles were mostly affected. Blood or tissue cultures were obtained in 94 (92%) and 59 (57%) patients, respectively. Of those, 55 (58%) blood and 52 (88%) tissue cultures were positive, mainly for Staphylococcus aureus (35 and 19, respectively) and Streptococcus pyogenes (12 and 15, respectively). All patients received antibiotic treatment during hospitalization for a median of 10 days (interquartile range: 7–17), followed by outpatient treatment for a further median of 16 days (interquartile range: 11–22) in 95 (93%) patients. Fifty-nine (57%) patients required surgery. Conclusions: Pyomyositis is a challenging diagnosis that requires a high level of awareness. Blood and/or tissue cultures revealed S. aureus and S. pyogenes as the predominant causative agents.


Definitions
We created a case definition to discriminate pyomyositis from bacterial myositis and pyomyositis associated with osteomyelitis (OM) or septic arthritis (SA), classified as follows: 1. Pyomyositis: compatible clinical presentation with radiographic detection of abscess formation, with or without bacterial pathogen detected by blood or tissue culture (s). 2. Bacterial myositis: compatible clinical presentation with radiographic signs of myositis but without abscess, with bacterial pathogen detected by blood or tissue culture (s) and without an alternative diagnosis.3. Pyomyositis associated with OM or SA: criteria for pyomyositis fulfilled in the presence of OM or SA.
Fever was defined as body temperature ≥38.0 °C.Tissue cultures (TC) were defined as cultures from punctures, intraoperative biopsies and/or, in the case of Kingella kingae, positive polymerase chain reaction (PCR) from throat swabs. 25

Statistical Analysis
Baseline data were summarized using descriptive statistics stratified by classification of pyomyositis.Analysis was conducted in R ( https://www.R-project.org/) and figures were produced using the package ggplot2 (https://ggplot2.tidyverse.org).Due to limited patient numbers, we did not perform formal statistical modeling on subgroups.

Patient Characteristics
A total of 331 patient records were identified, of which 102 patients (31%) from 10 hospitals were included in the final analysis (Fig. 1).Search for cases in one cantonal hospital remained without result.Of the 229 excluded patients most had an alternative diagnosis (eg, viral myositis) and failed to fulfill the pyomyositis and bacterial myositis case definitions.
Patient age ranged from 2 weeks to 17 years and included 3 neonates (2 with associated, 1 with bacterial myositis), the median age was 8 years [interquartile range (IQR): 3.9-12.6](Table 1).Both sexes were affected equally.The majority (87%) of patients did not have an underlying disease or immunosuppressive treatment.Within 4 weeks prior to hospitalization, one-third of the children had a reported trauma (usually at the site of consequent pyomyositis).Fifty-seven (56%) had a preceding viral infection, including 4 with varicella.The median duration of symptoms prior to admission was 3 days (IQR: 2-6).

Clinical Presentation
Most patients (97%) presented with a triad of fever, localized pain and impaired movement (Table 1).Further, local swelling was documented in the region of the affected muscle in 39 (38%) of patients.The swelling was observed more often with pyomyositis in the lower (20/28) and upper extremities (11/15) compared to the pelvic muscles (6/58).The pelvis and proximal lower extremity were the most affected localizations.Overall, the most involved muscles were M. iliopsoas (n = 20), M. obturatorius internus (n = 16) and M. quadriceps (n = 16) and while younger children were more likely to have pyomyositis in their extremities, pelvic muscle involvement was more frequent in adolescents (see Figure, Supplemental Digital Content 1, http://links.lww.com/INF/F544).In 36% of patients pyomyositis was multifocal.They did not differ from patients with unifocal pyomyositis regarding culture positivity rate, pathogen or patient age (data not shown).
Blood culture positivity rate increased with patient age from 28% in children ≤4 years to 87% in those ≥11 years (Table 3).The spectrum of isolated pathogens also depended on patient age with more K. kingae and S. pyogenes infections in those ≤4 years (29% and 33% of all isolates, respectively) compared to those ≥5 years (0% and 19%, respectively).Of note, 3 of the 6 K. kingae cases were only PCR positive in throat swabs and should conservatively be considered as presumptive cases.In contrast, S. aureus infections predominated in older age groups (53% in those ≥5 years versus 24% in ≤4 years).
C-reactive protein, erythrocyte sedimentation rate and white blood cell count median values were higher in patients with OM and SA-associated pyomyositis compared to those with pyomyositis alone (see Figures, Supplemental Digital Content 2-4, http:// links.lww.com/INF/F545).Differences regarding laboratory values were also noticed depending on the age of the patients and the isolated pathogen (Table 3).

Diagnostic Radiologic Imaging
Radiologic exams were performed in 96 (94%) of 102 patients on admission.In 47 (49%) of these 96 patients, the findings were compatible with bacterial myositis or pyomyositis (muscle changes or muscle abscess, respectively), with different detection rates depending on the modality (Table 4).If pyomyositis was diagnosed in the initial imaging study (n = 47) (Fig. 2), further radiological exams during hospitalization (n = 51) were mainly follow-up images (n = 41) or image-guided interventions (n = 7).If bacterial myositis or pyomyositis was not diagnosed on admission (n = 49), 95 additional images were performed during hospitalization.In 46 of these 49 patients, magnetic resonance imaging (MRI) was applied next.One case was not diagnosed during the initial admission but only on rehospitalization.In the remaining 6 patients with no imaging study on admission, a radiologic exam was performed later during their hospitalization; pyomyositis was diagnosed by MRI in 4 patients, by computed tomography (CT) in one patient and remained unproven in one patient.

Surgical Intervention
In 59 (58%) patients, surgical treatment was performed, including image-guided percutaneous abscess puncture (n = 30) and open surgical abscess drainage (n = 21) (Table 1).Seventeen patients received an intervention that was not related to the pyomyositis, primarily joint punctures (n = 13) in patients classified as SA associated with pyomyositis.

Special Patient Populations
Of the 13 patients (median age 13.5 years) with an underlying disease, 6 had a malignancy, that is leukemia (n = 3), lymphoma (n = 2) and Ewing sarcoma (n = 1).Five of them were receiving chemotherapy when pyomyositis (n = 4) and bacterial myositis (n = 1) were diagnosed and one patient with OM-associated pyomyositis was newly diagnosed with leukemia at the same time.These patients had a more severe clinical course with 4 of them having multifocal abscesses in up to 10 different muscles.The pathogens were Escherichia coli in 3, S. aureus in 2 and Group C β-hemolytic streptococcus in one patient.
Five patients initially presented with acute abdomen and were radiologically suspected to have appendicitis; 4 of them underwent laparoscopic appendectomy which rejected the diagnosis.

DISCUSSION
This study is, to our knowledge, the largest series of patients with bacterial myositis or pyomyositis in a nontropical country, comprising 102 patients in the pediatric age group.We found that this disease affects mainly healthy children of all ages, most commonly muscles of the lower extremities and the pelvis, with S. aureus and S. pyogenes as the main pathogens cultured, and that all patients received antibiotic treatment, although of various regimens and duration.
Similar to other studies, 3,[14][15][16] we observed an increase in pyomyositis cases over the time period but this was not linked to the emergence of resistant bacterial infections such as communityacquired methicillin-resistant S. aureus infections, described by Pannaraj et al 9 in the United States.It is possible that the increase is due to improved diagnostic procedures, specifically the broader use of MRI, and greater awareness among healthcare professionals.In addition, changes in environmental and lifestyle factors as well as global warming could also be contributing to the fact that the disease, which was once considered a tropical disease, is increasingly prevalent in temperate countries. 3,6,7,14,19We did not observe seasonal differences as described for rainy seasons in tropical countries, 6,21,26,27 which is attributed to the facilitated bacterial skin colonization in humid environments. 7,14yomyositis is most common in the first two decades of life. 1 Although most series report male predominance with percentages between 60% and 80%, 3,4,8,12 we did not observe this in our cohort.About 31% of patients had reported trauma in the previous 4 weeks, mainly affecting the same limb, and 56% had a viral infection.This is in accordance with other reports that identified trauma and other muscle injuries such as rigorous exercise, intramuscular injections and underlying viral myositis 1,5,6,8,21,28 as predisposing factors.Skeletal muscle, once damaged, becomes susceptible to hematogenous bacterial invasion. 11Our patient population consisted of mainly healthy children.Among the 13% with a chronic underlying disease, those 6 with a malignant disease differed most impressively from others, as they were more severely ill and often presented with disseminated disease.
Like other reports, [4][5][6] unspecific signs and symptoms such as fever, pain and impaired movement were documented in all our patients.Specific signs, such as swelling of the affected area, were lacking in 62% of the patients, especially in those with pyomyositis affecting deep muscles (6 of 58 patients ie, 10% with pelvic pyomyositis had swelling).Pyomyositis can present in 3 stages: (1) prodromal primary inflammation of the muscle, followed by (2) abscess formation, characterizing "pyomyositis," potentially with progression to (3) systemic manifestations and sepsis, if left untreated. 6,21Most patients (74%) included in this study had stage 2 disease (12% stage 1, 14% stage 3).
In only 20% of our patients was pyomyositis the suspected diagnosis on admission.Rather SA and/or OM were assumed and pyomyositis was usually found by means of MRI and other imaging studies.Of note, 25% still had pyomyositis in association with SA or OM.
In 5 of our patients, appendicitis was suspected and 4 of them had an appendectomy without confirming the diagnosis.Persistence of inflammatory signs prompted further differential diagnostic considerations and tests leading to a diagnosis of pyomyositis.

Characteristics of Pediatric Pyomyositis
helpful in determining targeted antibiotic treatment.We found a higher positivity rate of blood cultures in older children and adolescents.As the time from onset of symptoms to presentation was comparable across all age groups, this difference could possibly be explained by smaller blood volumes taken from children ≤4 years of age.Moreover, K. kingae infections are more difficult to detect by blood cultures and are more common in young children.There is one published case of K. kingae causing pyomyositis. 29Our study adds 3 more such cases plus 1 bacterial myositis and 2 cases of pyomyositis associated with SA.
Cultures of the abscess or surrounding fluids resulted in a high positivity rate (88%).14][15][16][18][19][20][21][22][26][27][28]30,31 Panton-Valentin-Leukocidin (PVL) producing S. aureus 4,6,12 have been associated with more severe courses of the disease with larger abscesses compared to non-PVL producing organisms. 9Production of PVL was not systematically tested in S. aureus isolates from our retrospective study, therefore we cannot confirm this observation.Beyond S. aureus, multiple other bacteria have been implicated, such as Gram-positive (S. pyogenes and S. pneumoniae) and Gram-negative bacteria (E.coli, Klebsiella pneumoniae)28 with notable differences by age of patients.In our study, S. pyogenes was detected in 24% of pathogen-positive cases; in fact, it was the most common pathogen in children ≤4 years of age.
Early diagnosis and prompt antibiotic treatment have been suggested as crucial factors influencing favorable outcome, resulting in complete recovery without complications. 4,6,7Assigning the course of infection into stages 1-3 can be helpful to define treatment strategies. 28While stage 1 disease can be managed with appropriate antibiotics only, 12 large abscesses often require combined antimicrobial and surgical treatment.In our study, like many other reports, 4,6,8 surgical abscess drainage was required in over 50% of patients with pyomyositis.Regarding the choice of empiric antibiotic treatment, recommendations are lacking.However, a regimen comparable to treatment of SA or OM seems reasonable, 32 including antistaphylococcal and antistreptococcal coverage 8 with modifications depending on local epidemiology and resistance patterns. 9In general, the duration of therapy should depend on clinical improvement 9 and host factors such as immunosuppression. 15reatment for a total of 3-4 weeks seems adequate, ideally initially via intravenous route and switching to an oral antibiotic as soon as possible. 9,28However, antibiotic treatment was continued beyond 4 weeks in 56% of our patients.
Patients with pyomyositis showed a longer duration of hospitalization, required more frequent surgical interventions, and showed more complications than patients with bacterial myositis.Possibly, early diagnosis and therapy before the abscess occurs could prevent complications, but early diagnosis in the absence of a detectable abscess is extremely challenging.Reported case fatality ranges between 0 and 10%, 1,4,21,28 but in general, most patients have excellent recovery.In our study, no fatalities occurred.
A major strength of our study is the large number of cases in the pediatric age group, combined with a standardized approach to data extraction.There are also limitations.First, due to the retrospective design, only recorded information was available for analysis.Second, although we included a broad ICD-10 diagnosis definition, we might have missed cases because of coding errors.Finally, pathogenicity factors of isolates, for example, PVL production, could not be determined retrospectively.In conclusion, pyomyositis should be considered and investigated in any child or adolescent presenting with fever, acute muscle pain and movement restriction, with or without a history of muscle injury.MRI plays a pivotal role in the timely diagnosis and precise localization of the disease.All efforts should be made to obtain microbiological cultures to identify the causative pathogen.With timely and adequate antibiotic therapy and surgical intervention, if needed, most patients have a favorable clinical course.

FIGURE 1 .
FIGURE 1. Flow chart of the study population.

TABLE 1 .
Patient Characteristics, Clinical Presentation and Course of Disease

TABLE 2 .
Laboratory Values and Diagnosis Category by Pathogen © 2024 The Author(s).Published by Wolters Kluwer Health, Inc