Pediatric central nervous system (CNS) tumors are the most common solid tumors in childhood and the leading cause of cancer-related death, accounting for 15%–20% of deaths among children aged 0 to 14 years globally (1–4). According to the Central Brain Tumor Registry of the United States, the incidence of primary CNS tumors is 6.14 cases per 100,000 children and adolescents (1, 5, 6), with 58% of these cases classified as malignant.

In high-income countries, the 5-year survival for pediatric patients with cancer exceeds 80%, reflecting the impact of early diagnosis and access to specialized healthcare services (7). However, global disparities exist, including in Latin America where the prevalence of childhood cancer is 11%, which is lower than that of Asian and African countries (8). In Colombia, CNS cancer in children is a significant health burden. The reported prevalence is 22.1% and it is the second leading cause of cancer-related death, with 47.4% of cases involving solid tumors (1, 2, 7). Moreover, 7,769 cases of childhood cancer were reported between January 2020 and January 2021, including 876 new cases. In southwestern Colombia, where our institution is geographically located, the prevalence of childhood cancer is 575 (95% CI, 545–606) per 1,000,000 individuals under the age of 18 years (2).

Precise tumor diagnosis and staging play a crucial role in determining treatment and prognosis. The World Health Organization (WHO) classification, with its 2021 update, places a strong emphasis on molecular diagnostics to distinguish tumors affecting adults from those affecting children, as this differentiation may influence treatment strategies and disease progression (9). CNS tumors in children, which originate from the brain, spinal cord, and meninges, are predominantly of primary origin, with gliomas comprising half of these tumors in children and adolescents (10, 11).

Surgical resection is the preferred treatment for CNS tumors in children. In Colombia, surgery has been the main treatment for children with CNS tumors since 2020, with an adoption rate of 98.7%. The surgical approach is tailored to the patient’s clinical condition and imaging findings (3, 12), with a focus on tumor mass reduction, intracranial hypertension management, hydrocephalus relief, or obtaining tumor samples for accurate classification. Postoperative care in a pediatric intensive care unit (PICU) is essential for continuous monitoring and early detection of complications such as infections, seizures, endocrine disturbances, or stroke (12).

Outcome data, including postsurgical data, provide valuable insights into care, risks, and opportunities for improvement in patient care. Despite Colombia’s effort to conduct epidemiological surveillance of pediatric cancer, there is limited information on postoperative outcomes following tumor resection. This study aims to describe the clinical outcomes and most common complications during the postoperative period for pediatric patients admitted to a PICU between 2011 and 2021 in a resource-limited setting that manages complex pediatric cases in southwestern Colombia.

This study included 114 patients who were admitted to the PICU immediately after undergoing CNS tumor resection. There were 63 males (55.3%) and 51 females (44.7%), with a median age of 8 years (IQR, 4–12) at the time of the first surgery. The median time from symptom onset to diagnosis was 29.5 days (IQR, 8–121), and the median time from diagnosis to surgery was 2 days (IQR, 1–8).

Before surgery, 21.1% (n=24) of patients experienced intracranial hypertension, with 12.2% (n=14) requiring emergent ventriculostomy prior to tumor resection. Due to cerebral edema pre-surgery, 78.7% (n=89) of patients received treatment with dexamethasone, and 79.8% (n=91) continued corticosteroid therapy for up to 5 days postoperatively. Only 15.7% (n=18) of patients required corticosteroid administration beyond the initial 5 days after surgery.

Regarding surgical outcomes, all patients underwent primary surgery, with 49.1% (n=56) achieving total tumor resection and 28.1% (n=32) undergoing partial resection (less than 90% resection), as documented in the neurosurgeon’s surgical notes. Surgical outcome information was not available in the medical records of 26 patients (22.8%). Tumor localization was primarily infratentorial in 51.7% of cases (n=59). The most prevalent tumor type by histology was embryonal (25.4%, n=29), followed by low-grade glioma (20.1%, n=23). Tumors were graded according to the 2016 WHO CNS Tumor Classification (14) as outlined in Table 1 . During surgery, 19.3% (n=22) of patients required blood transfusions. Additionally, 50% (n=57) of patients had ventriculostomy placement, either concurrently with tumor resection or in the early postoperative period.

After surgery, 33 patients (28.9%) had an uneventful postoperative phase. In some cases, multiple complications occurred in a single patient. Regarding the therapeutic interventions performed during the immediate postoperative phase (<24 hours), 32.5% (n=37) of patients required anticonvulsants, 17.5% (n=20) received antibiotics, 7.89% (n=9) were administered vasopressors, and 6.24% (n=7) received hyperosmolar therapy. Throughout their PICU stay, 24.5% (n=28) of patients required invasive mechanical ventilation (IMV) with a median duration of 3 days (IQR, 2–9). Additional details are shown in Table 2 .

Table 3 outlines the clinical characteristics and complications during early and late postoperative periods, grouped according to their timing. In the initial 24 hours following surgery, complications included surgical bed hemorrhage (6.14%, n=7), intracranial hypertension (3.50%, n=4), and septic shock (0.87%, n=1). Regarding neurologic status at 24 hours postoperatively, 89 (78%) patients had a Glasgow Coma Scale (GCS) score between 14 and 15, 10 (8.8%) had a score between 9 and 13, and 3 (2.6%) scored below 9. In 12 patients (10.5%), the GCS score was not available due to sedation and IMV at that time. Within the first 48 hours, the most common complications were motor deficits (36.8% of patients, n=42), facial paralysis (12.2%, n=14), and sensory deficits (11.4%, n=13). Diabetes insipidus, cerebellar syndrome, seizures, and hypothalamic insufficiency occurred in fewer patients. Between 2 and 7 days postoperatively, motor deficits persisted in 30.7% of patients (n=35), with notable occurrences of facial paralysis and sensory deficits. CNS infection was reported in 1 patient (0.87%). After 7 days, the incidence of motor deficits decreased to 13.2% (n=15), but sensory deficits, facial paralysis, seizures, and endocrine disorders remained. During this period, 3 patients (2.63%) developed infectious complications.

During the study period, 3.5% of patients died (n=4) during PICU stay. Although the causes of death were multifactorial, they were primarily related to complications that developed during the postoperative period rather than to the underlying disease. The majority of these deaths were due to CNS infections, sepsis, respiratory failure, and brain death. The FSS changed in some patients between admission and discharge, with 13.2% (n=15) of patients exhibiting a 1-point increase, 7.01% (n=8) showing a 2-point increase, and 10.5% (n=12) experiencing an increase of more than 2 points. Additionally, 5.26% (n=6) required tracheostomy and 13.2% (n=15) required gastrostomy during the PICU stay due to functional impairment and residual morbidity resulting from the primary disease and/or the surgical intervention. Patients received adjuvant chemotherapy and/or radiotherapy as part of their oncologic treatment after discharge from the PICU, but data on these additional treatments were not collected.

Postoperative complications in children after CNS tumor resection are critical and require intensive monitoring and continuous follow-up as many factors contribute to recovery and prognosis, including the patient’s preoperative condition and the type of surgery (15, 16). To date, limited data are available on the postsurgical outcomes following CNS tumor resections in Colombia. Our study reports the surgical morbidity and mortality of pediatric patients treated in a PICU of a tertiary referral center in Cali, Colombia.

The distribution of age and sex of our patients aligned with previous research conducted by Ramírez et al. and with the VIGICANCER report published in 2024 for this patient population in Colombia (17), as well as with data from other countries (1, 2, 5, 6, 11). Similarly, the anatomical location and WHO histopathologic classification were consistent with previous reports indicating that among gliomas, astrocytomas, medulloblastomas, and ependymomas are the most prevalent histologic subtypes (6, 9–11). The median PICU stay was 4 days (IQR, 2–7), which was comparable to that of other conditions managed in the PICU and did not involve prolonged stays (18, 19).

Immediate postoperative complications included surgical site hemorrhage (6.14%), intracranial hypertension (3.5%), and septic shock (0.87%). Hemodynamic complications were infrequent, with 7.89% of patients requiring vasopressor support and 6.14% receiving hyperosmolar therapy. Notably, no patients required reoperation, indicating effective clinical management and monitoring by a multidisciplinary team of pediatric intensivists, neurosurgeons, anesthesiologists, and nurses specialized in the care of critically ill children.

Neurologic assessment at 24 hours postoperatively revealed that 78% of patients were alert and oriented (GCS >14). Only a small proportion of patients (13%) had a GCS score below 9 at 24 hours, likely due to orotracheal intubation and sedation in the early postoperative period. Despite an average anesthesia duration of 4 hours, 24.5% of patients required IMV with a median duration of 3 days, which is considered a relatively short period of ventilatory support. This need was likely due to the complexity of the procedure rather than to pulmonary complications. Of these patients, 5.2% required tracheostomy placement prior to their PICU discharge.

In our study, we found that neurologic and endocrinologic complications were the most prominent complications, consistent with the existing literature (20). Specifically, motor deficits affected 36.8% of patients within 48 hours postoperatively, followed by facial paralysis (12.2%), sensory deficits (11.4%), and cerebellar syndrome (2.63%). These findings resemble those reported by Zhang et al., who documented a motor deficit prevalence of 56.8% in pediatric patients admitted to the PICU following tumor resection (21). A small proportion of our patients experienced seizures following surgery: 2.63% within the first 48 hours and 3.5% in the late postoperative period. These findings are also consistent with the existing literature, in which the reported prevalence ranges from 1.1% to 10.5% and often correlates with substantial bleeding or electrolyte imbalances, resulting in prolonged ICU stays (20, 22–24). The prophylactic use of anticonvulsants likely contributed to the reduced frequency of this complication in our pediatric cohort.

Endocrine disorders included diabetes insipidus in 5.26% of patients and hypothalamic insufficiency in 2.63% of patients within 48 hours post-surgery. The lower prevalence of these endocrine disorders in our study compared to the reported 37.1% in other studies (25) could be attributed to the limited evaluation period during the PICU stay, which may have excluded delayed-onset conditions. Other studies have linked endocrine disorders with prolonged PICU stay and increased risk of neurologic morbidity (16, 22, 26).

Postoperative infections were relatively rare in our study. One patient developed a CNS infection (0.87%) on the fifth day post-surgery, while 2 patients experienced CNS infection (1.75%) and 1 had bacteremia (0.87%) with pneumonia and surgical site infection in the late postoperative period (>7 days). These findings are consistent with publications reporting a 2.5% prevalence of CNS infection in neurosurgical patients (15, 24). Despite only 17.5% of our patients receiving immediate postoperative antibiotics, 20% displayed clear signs of infection. Although guidelines caution against postoperative antibiotic use as they do not decrease the frequency of infections (27), our findings support the use of prophylactic antibiotics to reduce the risk of surgical site infections and meningitis associated with ventriculostomy, emergent craniotomies, and early reinterventions in specific contexts (16, 22, 28).

Based on the FSS as a measurement of long-term sequelae, 10.5% of patients had residual functional morbidity after discharge, defined as more than a 2-point difference between PICU admission and discharge scores (29). Additionally, 5.26% of patients required tracheostomy and 13.2% required gastrostomy before discharge. These advanced interventions were conducted in adherence to clinical practice guidelines and adapted to each patient’s specific clinical complexity (30). Despite the low overall long-term morbidity, we must not underestimate the impact of morbidity on the quality of life and survival of pediatric patients. This is supported by findings from previous studies and further evidenced by the need for gastrostomy and tracheostomy due to severe neurologic impairment in this specific patient population (31). Therefore, proactive consideration of early tracheostomy and gastrostomy placement is crucial to facilitate timely rehabilitation and improve overall outcomes.

The mortality rate for our patient cohort was 3.5% and was attributed to various causes, including CNS infections, sepsis, ventilatory failure, and brain death, consistent with expected complications in this vulnerable population. Mortality in high-income countries has shown a decreasing trend over recent decades due to advancements in therapeutic technologies within ICUs (24, 32, 33). However, in low- and middle-income countries like Colombia, mortality remains relatively unchanged (34). Furthermore, the observed mortality in our study was higher than that from other international studies but consistent with data from countries with similar epidemiologic profiles, highlighting challenges related to healthcare access and economic constraints faced by our population.