Comparison of endoscopic sinus sampling versus intracranial sampling for microbiological diagnosis of intracranial infection in children: a case series and literature review

Intracranial infection is often associated with contiguous sinus infection, with Streptococcus intermedius being the most common pathogen. Microbiological assessment is possible via sinus or intracranial sampling. While a sinus approach is minimally invasive, it is not clear whether this yields definitive microbiological diagnosis leading to optimized antimicrobial therapy and avoidance of intracranial surgery. A retrospective review of a prospectively collected electronic departmental database identified patients between 2019 and 2022. Further demographic and microbiological information was obtained from electronic patient records and laboratory management systems. Thirty-one patients were identified with intracranial subdural and/or epidural empyema and concurrent sinus involvement during the 3-year study period. The median age of onset was 10 years with a slight male predominance (55%). All patients had intracranial sampling with 15 patients undergoing sinus sampling in addition. Only 1 patient (7%) demonstrated identical organism(s) grown from both samples. Streptococcus intermedius was the most common pathogen in intracranial samples. Thirteen patients (42%) had mixed organisms from their intracranial cultures and 57% of samples undergoing bacterial PCR identified additional organisms, predominantly anaerobes. Sinus samples had a significant addition of nasal flora and Staphylococcus aureus which was rarely grown from intracranial samples. Of concern, 7/14 (50%) of sinus samples did not identify the main intracranial pathogen diagnosed on intracranial culture and additional PCR. Literature review identified 21 studies where sinus drainage was used to treat intracranial empyemas, with only 6 authors reporting concurrent microbiology results. This confirmed our cohort to be the largest comparative study in the current literature. No center has observed a greater than 50% concordance in microbiological diagnoses. Endoscopic sinus surgery may have therapeutic benefit, but it is not an appropriate approach for microbiological diagnosis in pediatric subdural empyemas. High rates of contaminating nasal flora can lead to misdiagnosis and inappropriate treatment. Routine addition of 16S rRNA PCR to intracranial samples is recommended.


Introduction
Bacterial sinusitis is a common disease in children with benign evolution and outcome for the majority affected [1].Five to 10% of all upper respiratory tract infections in childhood are complicated by sinusitis [2].Complications of acute pediatric rhinosinusitis, including intracranial infections, have an incidence of approximately 3 per million population per year [3].In general, the frontal sinuses are poorly developed until the age of 10 and thus, it is believed that intracranial complications of pediatric sinusitis are relatively rare and mostly affect older children [4].The rare suppurative complications secondary to impaired drainage of paranasal sinuses have potential to result in intracanal sequelae such as subdural empyemas, extradural abscesses, or cerebral abscesses [5].
The principles of management first rely on neurosurgical drainage of localized collections and systemic antibiotics (duration 4-8 weeks) which should ideally be managed in a specialist tertiary center with multidisciplinary supportt [6,9].In some studies, adjuvant management by an otolaryngologist either in the form of an endoscopic sinus surgery has been observed to reduce the risk of abscess and the need for revision surgery [7][8][9].Causative organisms of sinusitis are predominantly aerobic and anaerobic streptococci, obligate anaerobes, Enterobacterales, Staphylococcus aureus, and Haemophilus spp.[10] Empiric antibiotic treatment regimens need to penetrate the central nervous system and cover likely aerobic and anaerobic bacteria.Beta-lactam based therapy is recommended with common regimens including third-generation cephalosporins with the addition of metronidazole or carbapenems.Therapy can subsequently be targeted to specific antimicrobial susceptibilities of identified pathogens [11].
The more classical pathophysiologic route of intracranial bacterial invasion is via the valveless diploic venous system that connects the intracranial venous system and the sinus mucosa vasculature [5].Focal osteitis of the posterior table is thought to be followed by a septic retrograde thrombophlebitis.Subdural empyema subsequently results from involvement of the subdural space.Osteitis of the posterior table may also produce an epidural abscess when purulent material collects between the bone and the dura.This can rupture creating a subdural empyema or may spread hematogenously to other intracranial spaces.Similarly, osteitis of other sinus walls may cause either a periorbital abscess or a pericranial abscess (i.e., Pott's puffy tumor).
Poor prognostic factors for subdural empyemas are severe encephalopathy at presentation, young age, and delay before initiation of appropriate therapy [12].Furthermore, despite improvements in survival rates (overall > 90%), 15-55% of children have permanent neurologic deficits such as seizures or hemiparesis [13].While functional endoscopic sinus surgery (FESS) is often part of the management to produce source control and identify the offending organism, it is our center's experience that concordance rates, between the organisms cultured in sinus and intracranial locations, are low.Here, we describe our quaternary center's experience analyzing the microbiological results of intracranial and FESS cultures.

Methods
Retrospective review of a prospectively collected electronic departmental database identified patients between 2019 and 2022.Further demographic and microbiological information was obtained from electronic patient records and laboratory management systems.Identification of cultured organisms was undertaken by routine bacteriology identification methodology (UK, Standards for Microbiology Investigations) using MALDI ToF (Bruker Diagnostics).PCR methods utilized both broad range and specific pathogen PCRs.Broad range 16S rRNA PCR was performed using in-house primers targeting the V1-3 and V4-5 regions of the 16S genome.Microbiological result concordance was defined as either full, partial, or no concordance.Full concordance results from the same organisms identified in both culture and PCR of intracranial sample compared to sinus sample.Partial concordance was defined as any organism that was identified in sinus sample corresponding to organisms identified in the intracranial sample by either culture or PCR.If the sinus sample did not identify any organism shown in the intracranial sample, then this was classified as no concordance.Categorical variables were reported using counts and proportion continuous variables were reported as mean (SD).

Staphylococcus aureus
No

Results
Thirty-one patients were identified with intracranial subdural and/or epidural empyema and concurrent sinus involvement during the 3-year study period.The mean age of onset was 9.92 years (SD 3.87 years) with a slight male predominance (55%).The mean time from diagnosis to operation was 1.71 days (SD 2.64 days).87.1% of surgeries were physically adjacent to the sinus infection, with 61% undergoing craniotomy and empyema washout and the remaining cohort undergoing burr hole aspiration of suppurative collection (Table 1).
All patients had intracranial sampling with 15 patients undergoing sinus sampling (Table 2).The most commonly isolated pathogen was Streptococcus intermedius found in 19 (61%) of patients (Fig. 1).Obligate anaerobic organisms made up the second most common group of organisms found by either culture or molecular methods (Fig. 2).Of the anaerobic bacteria, the most common species were Prevotella (8 patients), Fusobacterium (3 patients), and Parvimonas (3 patients).Thirteen patients (42%) had mixed organisms from their intracranial samples either identified by culture or PCR (Table 3).Twenty-one patients had an intracranial sample tested by PCR of which 57% (12 patients) had additional organism identified by PCR methods that were not identified on culture.Nineteen additional organisms were identified in the 12 patients, predominantly anaerobes (Fig. 4).In contrast to intracranial samples, sinus samples had a significant addition of nasal commensal flora with an increased isolation of Staphylococcus aureus, coagulase-negative staphylococci, and Corynebacterium species (Fig. 1).One patient had sinus surgery but no sample taken leaving 14 patients for analysis.Concerningly, 7/14 (50%) of sinus samples failed to identify the main pathogen (Streptococcus intermedius) compared to intracranial culture.Sinus samples rarely grew anaerobes and the addition of PCR substantially increased diagnostic yield.Only 1 patient (7%) demonstrated complete concordance of cultured organisms between intracranial and sinus samples.(Figs. 3 and 4, Table 3).Five patients had partial concordance and 8 patients had no concordance of microbiological results.Patients who grew S. aureus in their FESS sample (n = 6) had additional anti-staphylococcal antibiotic therapy with rifampicin (n = 2), vancomycin (n = 1), or linezolid therapy (n = 3) despite no evidence of S. aureus in intracranial samples via culture or PCR.(Tables 3 and 4).Of the 31 children commenced on antibiotics 0-8 days prior to neurosurgical acquisition of intracranial culture sample, all of the antibiotic regimes were revised on initial management in our center with further revision in view of the intracranial culture results in 16/31 children.Identification of S. aureus in the intracranial culture or PCR sample in 2 patients were  definitively managed with meropenem (Tables 3 and 4).All of the children were maintained on a course of definitive antibiotics for at least 6 weeks.Literature review identified 21 studies where sinus drainage was used to treat subdural empyema, with only 6 authors reporting concurrent microbiology results (Table 4).Literature review confirmed our cohort to be the largest comparative study and no center has observed a greater than 50% concordance in microbiological diagnoses (Table 5).

Discussion
Multiple studies have investigated the bacteriology of intracranial complications of sinusitis, and several patterns of growth have been described: anaerobic and aerobic Streptococcus species [14,15], Streptococcus pneumoniae [16], anaerobes [17], polymicrobial [18], Staphylococcus [19] infections.Recent series have highlighted the Streptococcus anginosus group as the most common organisms in intracranial complications [20][21][22].The group consists of the following 3 species: S. anginosus, S. constellatus, and S. intermedius [20], which are commensal organisms of the oral cavity [23].This is reflected in our data which show Streptococcus intermedius as the most common organism isolated from intracranial samples.Our experience of utilizing PCR techniques to enhance the diagnostic outcome shows the high incidence of mixed infection.There is evidence to suggest that mixed organisms contribute to pathogenesis in abscess formation, in particular anaerobes, allowing enhanced growth within an infected site [24][25][26].
The human nasopharynx is colonized with a plethora of bacterial organisms making up the normal flora that varies between individuals.The causative pathogen for subdural and/or extradural empyemas from sinusitis will originate from the sinuses but it can be diagnostically difficult to obtain the correct pathogen when mixed with normal commensals.We have shown clearly that sinus samples do not correlate reliably with the organisms found in the subsequent intracranial sample.More worryingly, in 50% of patients, the sinus samples did not identify the causative pathogen.The isolation of alternative organisms can lead to misdiagnosis and inappropriate antimicrobial therapy being administered.In our cohort, 4 patients were exposed to antibiotics (rifampicin, vancomycin, linezolid) to target S. aureus and 2 patients were definitively treated with linezolid.The isolation of S. aureus was not confirmed on intracranial sampling casting doubt on clinical significance and exposing these children to potentially more toxic microbial therapy.Streptococcus anginosus group and anaerobes are challenging organisms to culture in the routine laboratory; therefore, the diagnostic yield from sinus samples may be lower.We showed that a large proportion of patients have polymicrobial infection and the addition of routine PCR techniques increased our diagnostic yield by 57%.
The limitations of our study include the lack of comparator group who did not undergo neurosurgical intervention.Our group of patients are biased towards larger more complicated empyemas as these are the patients who would be transferred to our neurosurgical center.We therefore did not have a group that underwent FESS only and conservative treatment with antibiotics.However, as our paper primarily focuses on the comparative value of both techniques rather than clinical outcomes, this could be a consideration for a future study.It was also a single-center study and therefore potentially influenced by either surgical or microbiological laboratory practices that might not be generalizable to other centers.Our microbiology results do align with expected results from the literature review.

Conclusion
Our study is the largest cohort of concurrent microbiological sampling of sinus and intracranial sites in the literature for management of patients with intracranial suppurative infection in children.Endoscopic sinus surgery may have therapeutic benefit; however, it is not an appropriate approach for microbiological diagnosis in pediatric intracranial suppurative infection.High rates of contaminating nasal flora can lead to misdiagnosis and inappropriate treatment.Routine addition of 16S rRNA PCR to intracranial samples is recommended to increase diagnostic yield.

Fig. 1 Fig. 2
Fig. 1 Number of patients compared to the isolated organisms either via intracranial culture/PCR or sinus culture

Table 1
Patient demographics of concomitant sinusitis with intracranial infection requiring neurosurgical intervention

Table 2
Details of operative patient cohort whom underwent both intracranial and FESS microbiological sampling

Table 3
Additional microorganisms identified on the intracranial PCR analysis compared to the intracranial and FESS culture

Table 4
Empirical and definitive antibiotics provided

Table 5
Literature review sinus drainage in pediatric sinusitis with intracranial complications Staph aureus (sinus) and Peptococcus magnus and P. acnes