Introduction

Significant variation exists between the incidence of primary central nervous system tumors between high- and low- to middle-income countries (LMICs).1 More than 90% of children at risk of developing childhood cancer each year live in LMICs.2 Outcomes of these children are relatively good in high-income countries (HICs); however, in LMICs, outcomes are notably less optimistic.

For pediatric brain tumors, surgical resection remains the primary step of treatment.3 However, neurosurgery is associated with morbidity and mortality due to surgical site infections (SSIs).

The objectives of this study are to conduct an internal quality control, and to compare our findings with international outcomes. We aim to establish benchmark parameters for institutional practices that can be adopted by smaller, less specialized centers in the field.

Methods

This study involved retrospective review of patients' charts, diagnosed and treated for brain tumor, 0 to 18 years, from January 2011 till December 2022. Magnetic resonance imaging brain with contrast was used for initial confirmation of diagnosis, with a standardized formula of length x width x thickness x 0.523 to calculate tumor volumes.4 Those who did not opt for surgery were removed from the analysis list. Tumors were classified according to World Health Organization (WHO) classification of tumors.5

All patients received postoperative dexamethasone to manage peritumoral edema and perioperative parenchymal swelling.6 Resection status was described following the guidelines provided by Karschnia et al as follows: Gross total resection (GTR; removal of 100% of the tumor), near total resection (NTR; removal of >95% of the tumor), subtotal resection (STR; removal of ≥80% of the tumor), partial resection (PR; removal of 1-79% of the tumor), and biopsy (performed solely for diagnostic purposes).7

Files were systematically reviewed, encompassing postoperative notes, daily progress notes, and lab results, to identify patients who developed SSI. SSI was defined as any infection occurring near or at the incision site and/or deeper underlying tissue spaces and organs within 30 days of a surgical procedure.8 SSIs included suppuration of wounds, scalp abscess, meningitis, ventriculitis, and intracerebral abscess. Meningitis was diagnosed on the basis of cerebrospinal fluid (CSF) detailed report and culture and sensitivity (CS).9 Patients with deep infections were initially treated empirically with ceftriaxone and vancomycin, with adjustment of antibiotics guided by CSF CS reports. In case of deterioration in the patient's condition, an escalation of antibiotics to include meropenem, vancomycin, and colistin was implemented. Criteria of stopping antibiotics were total 14 days of treatment, with at least 10 days post-negative CSF and blood cultures, given that the patient achieved afebrile status and there was observable clinical improvement.10

Qualitative data included gender, residence, site of tumor, completion of treatment; presence, type, site of SSI, sensitivity of organisms, type of treatment received, prophylactic antibiotics received, acute postsurgical complications, and long-term sequalae. Continuous data included age, height, weight at diagnosis, volume of tumor, duration between symptoms onset and surgery, length of surgery, length of hospital stay, and dose of radiation therapy. Last date of follow-up was used to calculate survival time. Progression was confirmed radiologically.

Statistical Analysis

Since our data was not normally distributed, median and frequencies were calculated for quantitative and qualitative variables, respectively. Pearson chi-squared test was used to calculate correlation between qualitative variables. Independent sample t-test was used to calculate difference between means of two groups. Regression (linear and logistic) analysis was performed on variables that showed significant (p ≤ 0.05) correlation. Kaplan–Meier survival curves were used to calculate survival analysis. Log-rank test was used to compare survival between two groups. A p-value of less than or equal to 0.05 was considered significant. All data was analyzed using Statistical Package for the Social Sciences (SPSS) v22.

Results

From 2011 to 2022, 339 pediatric patients (202 males, 134 females) were diagnosed with brain tumors, with a median age of 108 (IQR: 170, 60) months (9 years) and a median duration between symptom onset and surgery of 60 (IQR: 180, 21) days (Table 1). Out of the total, 279 patients (83%) underwent neurosurgery and were included in the study.

Cerebellum was the commonest site of tumor presentation (101/339; 30%; Fig. 1). Median primary tumor volume was 36.85 (IQR: 86.5, 20) mL. Metastatic disease was present in 20/279 patients (7%). Hydrocephalus was present in half of the patients (166/279; 59%). Most frequent pathological diagnosis was circumscribed astrocytic glioma (74/279; 26%) (Table 2). Cancer predisposition syndrome testing was conducted in 74 out of 279 patients, with 30 testing positive (p53 mutation 29, constitutional mismatch repair deficiency [CMMRD] 1); the majority (73.3%) of positive cases were identified in diffuse high-grade gliomas (Table 3).

Close

Fig. 1: Distribution of tumor according to site. ATRT, atypical teratoid rhabdoid tumor; CNS, central nervous system; NOS, not otherwise specified.

Export to PPT

All patients received presurgery prophylactic antibiotics. Majority underwent elective surgery (255/279; 91%) (Table 2), with a median surgery duration of 4 (interquartile range [IQR]: 5.5, 2.5) hours. Craniotomy was the predominant procedure (275/279; 99%), with the commonest surgical approach being occipital (144/279; 52%). Resection status was GTR (129/279; 46%), STR (59/279; 21%), NTR (48/279; 17%), PR (20/279; 7%); while 32/279 (11%) patients underwent a biopsy only.

Majority with hydrocephalus (160/166; 96%) underwent a CSF diversion procedure (extraventricular drain [EVD] commonest 108/160; 67%), with majority (116/160; 72%) having CSF diversion at the time of tumor resection/biopsy. Thirty-nine (39/108; 36%) patients with EVD underwent subsequent ventriculoperitoneal shunting procedure, at a median time of 8 (range: 2–28) days.

Acute postsurgical complications developed in 107/279 (38%) patients, commonest being neurological (69/107; 64%). The incidence of other complications was; infections other than SSI (35/107; 33%), electrolytes imbalance (15/107; 14%), endocrinological (13/107; 12%), postoperative hematoma (10/107; 9%), CSF leak from wound site (10/107; 9%), and posterior fossa syndrome (10/107; 9%; Table 3).

The incidence of SSIs was 23/279 (8%; Table 3). Notably, all cases of SSI were deep-seated infections. Only 8/23 (35%) patients had a positive growth in CSF culture. The infective organisms identified were Staphylococcus aureus 2, Acinetobacter 2, Escherichia coli 1, Klebsiella pneumoniae 1, Aspergillus flavus 1, and Candida albicans 1. Average duration of hospitalization for patients who developed SSI was 20 days, in contrast to 8 days in those without SSI (P = 0.004).

Majority patients in our cohort (171/279; 61%) completed treatment for their cancer. Treatment abandonment was observed in 82/279 patients (29%), while 26/279 (9%) patients were unable to complete treatment due to postsurgical complications. Seventy-one (71/171; 41%) and (63/171; 37%) received chemotherapy and radiation therapy, respectively (Table 1, Figs. 2 and 3).

Close

Fig. 2: Chemotherapy status according to tumor pathology. ATRT, atypical teratoid rhabdoid tumor; CNS, central nervous system.

Export to PPT

Close

Fig. 3: Radiation therapy status according to tumor pathology. ATRT, atypical teratoid rhabdoid tumor; CNS, central nervous system.

Export to PPT

SSIs were significantly associated with presence of hydrocephalus (hazard ratio [HR]: 1.124; 95% confidence interval [CI]: 1.029–1.542; p = 0.050), performance of CSF diversion procedure (HR: 1.053; 95% CI: 1.007–1.400; p = 0.040), CSF leakage from wound site (HR: 1.886, 95% CI: 1.561–1.970, p = 0.002), increase in in-hospital stay of patients (HR: 1.143; 95% CI: 1.09–1.192; p = 0.000), delay in postoperative management (15/23; 65%) (HR: 1.983; 95% CI: 1.938–1.995; p = 0.010), decline in treatment completion (HR: 0.179; 95% CI: 0.837–0.915; p = 0.015), as well as poor survival (HR: 5.385; 95% CI: 2.181–13.294; p = 0.010); while there was inverse correlation between age of diagnosis and SSI (HR: 0.991; 95% CI: 0.984–0.999; p = 0.018; Table 4).

The overall survival of our cohort was 84.9% with a median follow-up time of 11 (IQR: 36, 1) months (Fig. 4). Survival was significantly lower (56.5%) in patients with SSI (p = 0.010; Fig. 5). Median time to death was 18 months. Primary disease progressed in (63/279;22%) patients. The progression-free survival was 77.4% with a median progression-free time of 8 (IQR: 28, 1) months (Fig. 6). Median time to progression was 9 (IQR: 24, 4.5) months.

Close

Fig. 4: Overall survival.

Export to PPT

Close

Fig. 5: Surgical site infection-dependent survival (p = 0.01).

Export to PPT

Close

Fig. 6: Progression-free survival.

Export to PPT

Discussion

SSIs represent a potentially preventable source of postoperative complications, mortality, and financial strain particularly in LMICs. The heightened occurrence of SSIs in LMICs is attributed to factors such as inadequate sterility, environmental pollution, compromised immunity, antibiotic resistance, and suboptimal vaccination coverage.11 The incidence reported from other centers in Pakistan is approximately 29 to 35%.12 Our study reports an SSI incidence of 8%.

Our data highlights that the performance of CSF diversion procedures is associated with an elevated risk of SSIs. Same has been identified in other studies.1314 This highlights the importance of avoiding CSF diversion procedures unless necessary. Implementation of a preshunt surgery checklist for infection control can effectively reduce the SSI rate attributable to this preventable cause, as reported by Lee et al.15

Inverse correlation was observed between age and the risk of SSI, aligning with the findings by Boethun et al.16 However, contrasting results have been reported by others. The conflicting findings highlight the need for further studies to establish a more definitive understanding of the relationship between age and the risk of SSI.

The incidence CSF leakage (3.7%) was higher in those undergoing infratentorial surgery compared to those with supratentorial surgeries (6 vs. 4) (Table 5); this, however, was not statistically significant (p = 0.434). Notably, no significance was found between craniotomy and craniectomy concerning CSF leakage, as the majority of our patients underwent craniotomy.

A significant association was observed between SSI and disease progression (p = 0.007). SSIs understandably result in delays in postsurgical management, including the administration of chemotherapy and radiation therapy. These delays ultimately contribute to disease progression.

An interesting finding in our study was the significant influence of the duration between symptom onset and surgery on the incidence of SSI. Notably, individuals with a longer duration of symptoms before surgery (229 vs. 80 days) exhibited a decreased incidence of SSIs (p = 0.000). High-grade tumors often prompt early presentation and the development of hydrocephalus.1718 Individuals who underwent CSF diversion had a shorter duration between symptom onset and surgery (127 vs. 339 days; p = 0.00). Similarly, those with poorer survival outcomes also demonstrated a shorter duration between symptoms and surgery (79 vs. 242 days; p = 0.00; Table 6).

Positive antimicrobial growth was observed in 35% (8/23) of our patients with SSI. Consistent with literature, mixed flora, encompassing both gram-positive and gram-negative organisms, was identified. Notably, fungal species growth was detected in two patients, a noteworthy finding as fungal growth is infrequently reported.

SSIs contribute to a spectrum of adverse effects, including prolonged hospital stays,19 which also elevates the risk of hospital-acquired infections, further prolonging the stay. In our study, the average extra length of stay due to SSI was 11.5 days (p = 0.004), representing a substantial wastage of healthcare resources.

Lastly, SSIs resulted in delays in postoperative care and a decline in the completion of treatment. These factors collectively had a detrimental impact on the overall survival of our cohort, with statistically significant differences observed in survival durations: (8.5 vs. 23.7 months; p = 0.00) and (5.98 vs. 33.60 months; p = 0.00), respectively.

Based on the above findings, we advocate strict adherence to the WHO guidelines for SSI prevention.20 These guidelines encompass the following key measures: I) Whenever possible, consider the use of oral/enteral nutrient-enhanced formulas for underweight patients to ensure adequate nutritional status before surgery. II) Patients should undergo a preoperative bath or shower, utilizing either plain soap or an antimicrobial soap. III) Consider perioperative intranasal applications of mupirocin 2% ointment, either alone or in combination with chlorhexidine gluconate (CHG) bodywash. IV) Avoid hair removal whenever possible. If necessary, use clippers instead of shaving, as shaving is strongly discouraged both preoperatively and in the operating room. V) Administer antibiotic prophylaxis within 120 minutes before incision. VI) Utilize alcohol-based antiseptic solutions containing CHG for surgical site skin preparation. VII) Refrain from using antimicrobial sealants following surgical site skin preparation. VIII) Minimize the utilization of CSF diversion procedures whenever feasible.

Conclusion

The occurrence of SSIs in our cohort mirrored that of HICs. Identified risk factors for SSIs included younger age, a shorter interval between symptom onset and surgery, the presence of hydrocephalus, and the performance of CSF diversion procedures.

Notably, SSIs in our study were linked to substantial additional costs and elevated mortality rates. Delays in postoperative management and incomplete treatment significantly contributed to these outcomes, underscoring the multifaceted impact of SSIs on healthcare resources and patient well-being.

Limitations and Strengths

Our study is retrospective in nature, and it may not fully address the gaps in patient records. Several confounding factors, such as variations in tumor pathology, could have potentially influenced the outcomes. Treatment protocols were not standardized across all patients, and surgeries were conducted by different surgeons with varying levels of expertise, potentially contributing to variations in SSIs. However, this study offers valuable insight into the causes and consequences of SSI in children with brain tumors, and serves as a catalyst for further endeavors aimed at reducing the occurrence of such infections.