Long-term treatment outcomes of pediatric low-grade gliomas treated at a university-based hospital

A multimodality approach is generally considered for pediatric low-grade gliomas (LGG); however, the optimal management remains uncertain. The objective of the study was to evaluate treatment outcomes of pediatric LGG, focusing on long-term survival and factors related to outcomes. A retrospective review of 77 pediatric LGG cases treated at Ramathibodi Hospital, Thailand between 2000 and 2018 was performed. The inclusion criteria were all pediatric LGG cases aged ≤ 15 years. Diffuse intrinsic pontine gliomas and spinal cord tumors were excluded. The median follow-up time was 8.2 years (range, 0.6–19.7). The median age at diagnosis was 6.2 years (interquartile range, 3.6–11.4). Treatments modality included tumor surgery (93%), chemotherapy (40%), and radiation therapy (14%). The 10-year overall survival (OS) and 10-year progression-free survival were 94% and 59%, respectively, for the entire cohort. The 10-year OS was 100% in three subgroups of patients: pilocytic subtype, WHO grade 1 tumors, and recipient of gross total resection. After multivariable analysis, no tumor surgery had a significantly unfavorable influence on overall survival. With a multimodality approach, pediatric LGGs had excellent outcome. Gross total resection is the standard primary treatment. Chemotherapy is the alternative standard treatment in incomplete resection cases, unresectable patients, or patients with progressive disease. Radiation therapy should be reserved as a salvage treatment option because of late complications that usually affect patients’ quality of life.


Introduction
Pediatric low-grade gliomas (LGG) are defined as World Health Organization (WHO) criteria grade 1 and 2 tumors with at least some glial cell lineage components, and they are the most frequent primary childhood central nervous system tumors [1][2][3][4]. The prognosis is excellent when complete tumor resection is achieved. However, complete resection is not feasible in all LGG patients, mainly because of tumor location. Chemotherapy and radiation therapy (RT) are used as adjuvant treatment after partial resection (PR) or primary treatment for unresectable tumors. Nevertheless, there is no standard guideline for the management of post-surgical residual or unresectable cases, which is generally based on individual institute experience. Currently, chemotherapy is commonly offered, whereas RT is only administered occasionally because of the concerning long-term complications, especially in younger patients [2,5]. The literature regarding long-term outcomes in pediatric LGG patients treated with a multimodality approach remains limited. This study aimed to evaluate the long-term treatment outcomes in pediatric LGG, focusing on long-term survival and factors related to outcomes at a university-based institute.

Patient population
Data on pediatric LGG patients treated at Ramathibodi Hospital, Bangkok, Thailand between January 2000 and January 2018 were retrospectively reviewed and analyzed. The inclusion criteria were a diagnosis of pediatric LGG and patient age of 15 years or younger at the time of diagnosis. All patients were diagnosed by histological confirmation of WHO grade 1 or 2 glial cell tumors, except for patients with unresectable tumors, who were diagnosed by magnetic resonance imaging (MRI). Patients diagnosed with diffuse intrinsic pontine gliomas and spinal cord tumors were excluded from the study. The study was approved by the Ethical Committee of the Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand.

Clinical characteristics
Age at diagnosis, sex, histological subtype, tumor location, and tumor size measure in two dimensions were included in the demographic information. Tumor grade was defined following the WHO grade classification. Details on surgical resection, chemotherapy, and RT as the first course of treatment were identified. Postoperative MRIs, within 72 h after surgery, were examined and the extent of resection was categorized into four groups: gross total resection, subtotal resection, partial resection and biopsy, as narrated by Gnekow et al. [2]. Patients without surgical intervention were labeled "no tumor surgery." The patient's age was divided into 3 categories, including ≤ 5, 6-10, and > 10 years.

Overall management policy
In each pediatric LGG case treated in our institute, maximal safe resection with complete tumor removal was performed when feasible. In general, patients who underwent complete resection did not receive any adjuvant treatment. For patients who received incomplete resection, whether to proceed with adjuvant treatment was discussed in a pediatric tumor board meeting. Although there are no standard guidelines for adjuvant treatment in post-surgical residual or unresectable cases, we created a multidisciplinary team and developed national protocols for treatment. On the basis of our protocol, chemotherapy is accepted as the standard treatment in incomplete resection cases, unresectable patients, or patients with progression and recurrence. RT is usually offered as salvage treatment after unsuccessful redo surgery or multicourse chemotherapy.

Radiation therapy
Radiation technique, treatment volume, and radiation dose were mainly based on physician preference, and the year in which the patient was treated. In the earlier period of the study, patients were treated via a 2-dimensional RT technique using a cobalt-60 machine (Theratron Elite 80; MDS Canada Inc., Kanata, Canada). Since 2007, we have used a 2100C linear accelerator (Varian Medical Systems, Palo Alto, CA, USA), which was later replaced with a Clinic IX machine (Varian Medical Systems) and the Edge Radiosurgery system (Varian Medical Systems). More advanced RT techniques such as three-dimensional conformal RT, intensity-modulated RT, and volume metric arc therapy were used for most patients. Local-field radiation with a total dose of 45-54 Gy and daily doses of 1.8-2 Gy/fraction was commonly used in the study.

Assessment of outcomes and complications
Clinical assessment combined with neurological examination and MRI was regularly performed to assess outcomes. Followup examinations were conducted every 3 months for the first 2 years, followed by every 6 months for 5 years and then annually. The information regarding long-term treatment-related outcomes, neurological status, endocrine function, hearing status, visual acuity, cerebral vasculopathy, and secondary tumors was identified. Imaging follow-up was performed every 3-6 months for the first 2 years and then annually. The Response Assessment in Pediatric Neuro-Oncology (RAPNO) criteria were used by neuroradiologists to evaluate the radiologic response [6]. According to the RAPNO definition, a complete response (CR) is when there is no radiological evidence of a tumor on an MRI scan, a partial response (PR) is when there is a 50% reduction in all detectable tumor, and a stable disease (SD) is when there is a 25% reduction in all tumor. A 25% or more increase in tumor size or a new tumor in a previously unaffected location was considered a progression disease (PD).

Statistical analysis
Categorical data were presented as frequencies and percentages, and continuous data were shown as the median and interquartile range (IQR) or range. Overall survival (OS) was calculated as the time from the date of diagnosis to the date of death or the last follow-up. Progression-free survival (PFS) was defined as the time from the date of first treatment to the date of the first documented intracranial progression/recurrence by RAPNO criteria, all-cause death, or last follow-up. Survival was calculated using the Kaplan-Meier method. The log-rank test was used to compare differences in survival curves in univariate analysis. Variables with p < 0.05 were considered statistically significant and further analyzed in the multivariate Cox proportional hazard model. Variables included age at diagnosis (both continuous and categorical: 1-4, 5-10, and > 10), gender, pilocytic astrocytoma (yes vs. no), WHO grade (grade 1 vs. grade 2), tumor location (cerebral hemisphere vs supratentotial midline vs. posterior fossa), surgical treatment (no tumor surgery vs. 4 categories of extent of resection), and treatment modalities. Statistical analyses were performed using STATA software (17.0, StataCorp, College Station, TX, USA).

Patient characteristics
From January 2000 to January 2018, 94 patients were assessed for eligibility. Seventeen patients were excluded from the analysis because of exclusion criteria (n = 11) and insufficient information (n = 6). Thus, 77 patients were included in the analysis. There were 46 (60%) boys and 31 (40%) girls with a median age at diagnosis of 6.2 years (IQR, 3.6-11.4 years). Pilocytic astrocytoma, WHO grade 1 (51%), and diffuse astrocytoma, WHO grade 2 (31%), were the two most common tumor subtypes. Nine patients (12%) presented low-grade astrocytic gliomas, not otherwise specified (LGG, NOS), including one with WHO grade 1 and six with WHO grade 2. In 5 (6%) patients, the diagnosis was based on imaging only. Forty patients (57%) had WHO grade 1 tumors, while 30 patients (43%) had WHO grade 2 tumors. Four patients (5%) were diagnosed with neurofibromatosis type 1. The median tumor size was 3.7 cm (IQR, 2.6-4.6 cm). The tumor location was the posterior fossa in 36 cases (47%), supratentorial midline in 30 cases (39%) and cerebral hemisphere in 11 cases (14%) The baseline patient and tumor characteristics were summarized in Table 1.

Treatment modalities at the first diagnosis
Among 77 patients, surgical resection was performed in 55 patients (71%). The extent of resection was GTR in 33 patients (43%), STR in 5 patients (6%), PR in 17 patients (22%), and biopsy in 17 patients (22%). Due to extensive diffuse infiltrative (n = 2) or optic pathway sites (n = 3) tumors, five patients (7%) underwent no surgical resection. No severe complications or deaths occurred during the perioperative period. No patients who received GTR underwent any adjuvant treatment, except for two patients in the GTR group who received adjuvant chemotherapy and RT because of a concern for microscopic residual disease because the lesion was located in the optic pathway and midline region. All 5 patient who received STR and 6 patients who received PR underwent observation even though they had a residual tumor. The reasons for observation included very young age, WHO grade 1 and family preference. The remaining of 35 patients were administered further treatment modalities, including chemotherapy in 29 patients, RT in 4 patients and chemoradiation in 2 patients. Figure 1 revealed the overview of study population.
The six patients receiving radiotherapy or chemoradiotherapy were all treated with a median dosage of 54 Gy (range, 39.6-54 Gy) via conventional fractionation. The radiation technique also varied over time and included 2-dimensional RT (36%), 3-dimensional conformal RT (45%), and intensity-modulated RT (18%). The baseline and treatment characteristics of 77 patients were summarized in Table 1.

Disease course after initial management
The median follow-up time was 8.2 years (range, 0.6-19.7 years). At the last follow-up, 71/77 patients were alive, including 36 patients with stable disease, 33 with complete responses, and two patients with recent disease progression who were still receiving chemotherapy. Overall, 29/77 patients (38%) developed tumor progression once (n = 9) or up to four times (n = 20) following surgery alone (n = 10, 34%), multimodalities treatment (n = 10, 34%), chemotherapy alone (n = 8, 29%), and RT alone (n = 1, 3%). The median time to first progression was 1.2 years (range, 0.1-6.9 years). Most cases of tumor progression were localized disease (96%), while only one patient (4%) who had intraventricular grade 2 glioma developed local recurrence with leptomeningeal spreading. All patients with progression had tumors that resembled those that were present prior to the initial resection, with the exception of one patient who initially had a grade 2 astrocytoma in the cerebellum that had had adjuvant chemoradiotherapy. Three years later, when the primary tumor grew larger, it needed to be removed. The pathology report made reference to glioblastoma multiforme. A patient with diffuse thalamic astrocytoma got definitive RT. The patient developed intracranial meningioma 15 years later. He is still alive and asymptomatic after the meningioma's removal. Figure 1 revealed the overview of study population.
Six patients died. Five deaths were related to tumor progression, including one after malignant transformation, and one patient with local tumor control died from an unrelated severe pneumonia infection.

Management after tumor progression
Tumor progression developed in 29 patients with a median time of 1.2 years (range, 0.1-6.9) after the first management. Additional surgery for tumor progression was performed in 13 patients (13/29; 45%). Nine patients received GTR, whereas four patients had a residual tumor after re-operation that was further treated with adjuvant chemotherapy in three patients and chemoradiation in one patient. Sixteen patients (16/29; 55%) did not receive re-operation because of a diffuse pattern, tumor adhesion to critical structures, clinical progression only, multiple metastases in the brain and spine, or family preference. Among these 16 patients, 12 received repeated courses of chemotherapy, three patients received chemoradiation, and only one patient received RT alone. The chemotherapy regimen for salvage treatment was vinblastine (43%) and various combinations of carboplatin, vincristine, irinotecan, or temozolomide. All patients received local RT as salvage treatment, except one patient with multiple recurrences in the brain and spine who received craniospinal irradiation. No patient received re-irradiation in this study.
In total, 13 of 29 patients required multiple courses of chemotherapy and multimodality treatment because of repeated tumor progression. Following multiple treatments, seven patients were alive with a stable primary tumor, one patient with primary progression and leptomeningeal spreading showed stable disease after craniospinal irradiation and repeated courses of chemotherapy, and five patients died as a result of an uncontrolled tumor.

Treatment-related complications
No deaths or permanent focal neurologic deficits were caused by tumor resection. The acute toxicities of chemotherapy differed depending on the type of chemotherapy. Grades 3-4 febrile neutropenia occurred more often with carboplatinbased regimens than with the vinblastine regimen, which frequently caused mild side effects. The acute chemotherapy side effects included grades 3-4 febrile neutropenia in nine patients, grade 2 transaminitis in two patients, and grade 4 allergic reaction in one patient. Late sequelae presumably related to RT included endocrine dysfunction (n = 2), morbid obesity (n = 2), azoospermia (n = 1), malignant transformation (n = 1), and secondary meningioma (n = 1).

Discussion
Developments in the advanced management of pediatric LGG, including surgery, RT, and systemic treatment, have led to improved outcomes. Excellent long-term survival after GTR has been reported in several studies. The risk for tumor recurrence in patients who undergo complete tumor resection is extremely low [7][8][9]. For patients with incomplete tumor resection or unresectable tumors, appropriate nonsurgical management can include observation, traditional chemotherapy, and RT. Information regarding the extent and type of initial surgical interventions and clinical symptoms of the patient is essential to decide the type of further adjuvant treatment. Tumor progression after incomplete resection occurs independent of tumor location, and observation is justified in patients without tumor-related symptoms and those with small residual tumors [2,7,9,10]. Delaying the start of non-surgical treatment, including chemotherapy or RT, may increase the risk of irreversible neurologic impairment and may be associated with a potentially lower probability of tumor control. Nevertheless, there are no apparent differences in long-term OS following immediate versus delayed RT or chemotherapy [11][12][13][14]. The value of RT in improving PFS is well-documented [15]. However, RT is associated with numerous long-term complications, such as neurocognitive impairment, endocrinopathy, vasculopathy, and secondary tumors, especially in children under 5 years of age [16][17][18]. Therefore, chemotherapy is accepted as the primary nonsurgical treatment for pediatric LGG patients regardless of age, histology, and site to improve PFS and delay the use of RT [19]. The use of upfront chemotherapy in residual or unresectable pediatric LGGs as a strategy to delay or avoid the neurotoxic effects of RT has been used in our institute for many years. Thus, the main objective of this study was to report the long-term outcomes of a relatively large number of pediatric LGG patients treated with a comprehensive strategy following the university-based institute protocol.
In the current study, the excellent OS rate of 94% at 10 years was comparable to the range in several previous studies that reported 5-to 10-year OS rates between 80 and 90% [20,21]. However, the PFS rates of 23-50% were lower than the OS rates [7,8,19,[22][23][24]. These findings suggested the importance of regular long-term follow-up, early detection of disease progression, and comprehensive salvage treatment. In our Fig. 4 Progression-free survival stratified by WHO grade study, among 12 patients who underwent partial tumor resection without non-surgical adjuvant treatment, nearly half (42%) developed tumor progression. However, successful salvage tumor resection was performed. This reflected the fact that in pediatric LGG tumor recurrence cases, the tumor remained well localized to the initial site without increased invasiveness. This pattern of progression made the lesion responsive to repeat resection followed by adjuvant treatment if indicated. Careful neurosurgical follow-up is necessary for successful salvage surgery.
Prognostic factors affecting survival varied between previous studies; however, widely accepted important factors included the extent of surgical resection, pilocytic subtype, WHO grade, and patient age [8,9,12,20,22]. Consistent with previous studies, the current study confirmed the excellent OS of 100% in three patient subgroups: pilocytic astrocytoma, WHO grade 1 tumor, and receipt of GTR.
Pilocytic astrocytomas have a better prognosis than most other non-pilocytic subtypes. After GTR, the OS and PFS rates reported in several studies were 75-100% [8,11,25]. In cases of PR without any adjuvant treatment, the 5-year PFS rate was 50-60% [9,26]. It is unclear whether this is because pilocytic tumors are usually well circumscribed and more responsive to GTR than non-pilocytic gliomas or whether they represent a more biologically favorable tumor [9].
Numerous previous studies [8,9] reported that a greater extent of resection was correlated with better OS and PFS; however, the extent of resection was not associated with a statistically significant improvement in OS in our study. Nevertheless, patients who received partial to gross total resection showed better survival. Maximum surgical resection not only achieved an excellent OS rate but also prolonged PFS following one or repeated tumor resections without requiring adjuvant treatment in 58% of patients in the present cohort. This finding emphasizes the value of surgery, particularly partial to gross total resection. No tumor surgery was found to be a significant predictive factor for OS in our study. This finding may be related to the clinical features of extensive infiltrative disease and an eloquent site that is inaccessible for surgery. However, due to a small sample size, it is challenging to draw a conclusion in this issue.
Patient age was previously reported to be associated with survival in various studies of pediatric LGGs in all locations; however, the prognostic impact of older or younger age in pediatric LGG remains inconclusive. Age younger than 1-2 years at diagnosis was reported as an unfavorable prognostic factor by some studies [7,8,19,20,22,23,27]. Moreover, age older than 8-12 years has been shown to be associated with better or impaired prognosis, depending on an unfavorable histology [22] or tumor site [19]. In our study, there was no correlation between age and OS. The inconsistent results might be due to the retrospective nature of studies or differences in baseline characteristics and treatment modalities, especially the use of adjuvant chemotherapy and RT, which usually depends on the experience at each institution.
Chemotherapy is accepted as the first-line non-surgical treatment in incomplete resection or unresectable pediatric LGG cases. Two randomized control trials [19,22] and many small studies [7,8,19,22] reported similar results for all single drugs and combinations with a high primary tumor control rate of 70-95%; however, none achieved sustainable long-term tumor control with a decline in PFS to below 50% at 5 years. The commonly used chemotherapy regimens in our center include vinblastine-and carboplatin-based regimens based on a previous study [22].
RT is typically applied as conventional fractionated external beam radiation with a total dose of 45-54 Gy. Various RT techniques have achieved high PFS rates of 65-80% and OS rates of 93-98% after 5-10 years [28][29][30][31][32]. The use of postoperative RT showed a PFS benefit in some studies; however, adjuvant RT did not translate to an OS benefit [26,32]. In addition, RT was associated with malignant transformation and late cognitive and endocrine dysfunction in irradiated patients [26,31,33]. Although the present study showed acceptable PFS rates with chemotherapy for residual LGG cases and reported late RT complications, including endocrinopathy and RT-induced secondary tumors, RT should be avoided and chemotherapy may be an option for unresectable symptomatic, progressing or recurrent LGGs in children. In some cases that might need RT, a multidisciplinary team approach with modern RT techniques is recommended.

Conclusion
With a multimodality approach, pediatric LGGs had great OS. Maximal safe resection with complete tumor removal should be performed whenever feasible. Chemotherapy may be the best option for first-line adjuvant treatment, especially in patients with extended tumors and young children. RT should be preserved as salvage treatment after maximum resection and multicourse chemotherapy. Patient-specific decision-making is required and should be discussed in a multidisciplinary setting.
Author contribution Apisak Jujui-eam: conceived and planned the study, collected data and analysis of data, contributed to the interpretation of the results, prepared figure and table, wrote manuscript, discussed the results, and contributed to the final manuscript; Nongnuch Sirachainan: conceived and planned the study, contributed to the interpretation of the results, wrote manuscript, discussed the results, and contributed to the final manuscript; Suradej Hongeng: conceived and planned the study, contributed to the interpretation of the results, wrote manuscript, discussed the results, and contributed to the final manuscript; Ake Hansasuta: conceived and planned the study, contributed to the interpretation of the results, wrote manuscript, discussed the results, and contributed to the final manuscript; Atthaporn Boongird: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Oranan Tritanon: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Mantana Dhanachai: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Thiti Swangsilpa: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Rawee Ruangkanchanasetr: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Rasin Worawongsakul: conceived and planned the study, contributed to the interpretation of the results, discussed the results, and contributed to the final manuscript; Putipun Puataweepong: conceived and planned the study, collected data and analysis of data, contributed to the interpretation of the results, prepared figure and table, wrote manuscript, discussed the results, and contributed to the final manuscript.

Availability of data and material
The data that support the finding of this study are available from the corresponding author upon reasonable request.