The role of radiotherapy and chemotherapy in adult optic nerve gliomas: a SEER-based analysis

Background: Optic nerve gliomas (ONGs) are uncommon tumors of the central nervous system in adults. The aim of this study was to dene their characteristics, prognostic factors, and the impacts of adjuvant radiotherapy (RT) and chemotherapy on outcomes. Methods: Adult patients (age ≥ 18 years) with ONGs from the Surveillance, Epidemiology, and End Results (SEER) database were included. Univariate and multivariate Cox regression models were utilized to analyze the factors associated with survival. Kaplan-Meier method was used to evaluate the impacts of adjuvant therapies on overall survival (OS). Results: A total of 179 adult patients diagnosed with ONGs were identied between 1991 and 2016, with a median follow-up period of 64.0 months. The median age at diagnosis was 41.0 years. After excluding 18 patients with unknown information, the remaining patients included 142 (88.2%) low-grade tumors and 19 (11.8%) high-grade tumors. Multivariate analysis showed age at diagnosis, tumor grade, adjuvant chemotherapy were signicant factors for OS. The 5-year OS rates for patients with low- and high-grade ONGs were 85.5% and 10.5%, respectively. The employment of adjuvant RT or chemotherapy would signicantly shorten OS time in the low-grade group and could not prolong OS time in the high-grade group. Conclusions: This is the largest retrospective study of adult ONGs up to date. The overall prognosis of high-grade ONGs in adult patients is still poor despite multi-modality treatments. Adjuvant RT or chemotherapy might not be considered in adult patients with low-grade ONGs unless the malignant transformation or aggressive progression has been conrmed.


Introduction
Optic nerve gliomas (ONGs) are relatively rare neoplasms, accounting for only 1% of all intracranial tumors and 2% -5% of childhood central nervous system tumors. [1][2][3][4] Histopathological types vary, ranging from the most common WHO grade I pilocytic astrocytoma to uncommon grade IV high-grade glioblastoma. [5,6] They are preferentially occurred in the pediatric population, particularly the rst decade, and the majority are low-grade lesions. [7] ONGs can either be sporadic [8], or in association with Neuro bromatosis-1 (NF-1) [9], and the typical symptoms include decreased visual acuity, optic nerve atrophy, unilateral visual loss, or proptosis. [10,11] The optional treatment of ONGs remains controversial, ranging from clinical observation, surgical resection, chemotherapy, and radiotherapy (RT), due to the unpredictable course. [12,1] The treatment is generally commenced when radiological tumor progression or clinical deterioration occurs, especially visual decline. [13] For children, chemotherapy has been the rst-line treatment over the past two decades because of fewer adverse side effects than occurs with RT. [14,15,4] While in adults, ONGs are rare, and the treatment for which has historically been approached in a different manner than children. [16][17][18] However, the de nitive e cacy of adjuvant RT or chemotherapy is still inconclusive due to the paucity of research data. Also, because ONGs are relatively uncommon in adults, it is still challenging to perform prospective studies now or soon.
In the current study, we aim to report the largest series of adult patients with ONGs based on the Surveillance, Epidemiology, and End Results (SEER) database. To better understand these rare tumors, our retrospective study investigated the epidemiology, prognostic factors, and in particular, the impact of adjuvant RT and chemotherapy on outcomes further to de ne the role of adjuvant therapies in adult ONGs.

Study Population
The data information for this study was obtained from the recent SEER database, which provides clinical incidence, treatment, and survival data on many tumors and covering nearly 36.7% of the US population according to the 2010 census, and is maintained by the National Cancer Institute, Division of Cancer and 2010 -2016. The tumor site was categorized into unilateral, bilateral, and unknown. The extent of surgical resection was divided into ve groups, including gross total resection (GTR), subtotal resection (STR), biopsy, no surgery, and unknown on the basis of SEER surgery codes guidelines and another previous study. [19] For further survival analysis, patients with the unknown grade, unknown survival time, unknown surgery (code 90 and 99), and unknown radiation record were excluded, and we categorized age as "< 50 years" and "≥ 50 years", according to the previous literature. [19] Baseline patient characteristics were summarized by standard descriptive statistics and frequency tabulation. Overall survival analysis was measured by using the Kaplan-Meier method and compared with the log-rank test. Univariate and multivariate Cox regression models were utilized to assess the effect of variables of interest on OS. All statistical analyses were carried out in SPSS software version 25.0 (IBM Corp., Armonk, NY, USA), and the statistically signi cant standard was P < 0.05. Ethical approval or informed consent was waivered for this study because of the de-identi ed information of the patients included in the SEER.

Clinical Characteristics
A total of 179 adult patients diagnosed with ONGs were identi ed between 1991 and 2016. No patients with ONGs were recorded in the SEER database prior to 1991. Of the whole population, the mean and median ages at diagnosis were 42.1 and 41.0 years (range, 18 -101 years), respectively, with 55.9% of patients being female, 78.2% white people, and 43.6% of married status (Table 1). 116 patients (64.8%) were located unilaterally regarding the tumor site, and 12 patients (6.7%) were bilateral. For tumor grade, 81.6% had a low-grade tumor, and 11.2% had a high-grade tumor. After excluding 18 patients with unknown information, the remaining patients included 142 (88.2%) low-grade tumors and 19 (11.8%) high-grade tumors. In the low-grade tumor group, 78.2% of patients were less than 50 years; however, 89.5% of patients were more than 50 years in the high-grade group.  Due to the possible negative impact of the adjuvant therapies on OS, further analyses were performed in the subgroups of low-grade and high-grade tumors to clarify better the prognosis of these patients who received adjuvant therapies. For the low-grade tumor group, the 5-year OS rate was 85.5%; however, the 5year OS rate for patients in the high-grade group was only 10.5% (P < 0.001, log-rank test; Fig. 1). Based on the low-grade subgroup analysis, patients with age less than 50 years, no adjuvant RT or chemotherapy had the best OS time, and the difference was signi cant (P < 0.05, log-rank test; Fig. 2A-C).
The results of multivariate Cox regression also showed that younger age, and no adjuvant chemotherapy were signi cant factors for longer OS (Table 3). However, different results were found in the other subgroup of the high-grade tumor, which demonstrated that only older age resulted in a signi cantly shorter OS (P < 0.05, log-rank test; Fig. 2D). Moreover, there were no signi cant differences in OS for highgrade patients who received the adjuvant therapies or not, according to multivariate Cox regression ( Table  4, Fig. 2E-F). cases on the basis of the SEER database, is the largest series of adult ONGs and is the only study to compare the effects of different treatment strategies (including surgery, adjuvant RT, and chemotherapy) on survival with a large number of adult patients. Furthermore, our data identify that both adjuvant treatments (RT and chemotherapy) have a negative impact on the survival and prognosis of adult patients with low-grade tumors.

Epidemiological and Tumor Characteristics
ONGs in adults are extremely uncommon and were rst described by Hoyt et al. in 1973, which included 15 adult patients with malignant optic glioma. [25] In our study, 70% of the patients were below the age of 50, especially in the low-grade tumors, and the mean age at diagnosis was 42.1 years, which is nearly consistent with the previously limited reported mean age for patients diagnosed at the adulthood of 39 years. [17] However, patients diagnosed with high-grade ONGs usually were older than those with lowgrade ONGs, ranging from 57 to 66 years, [23,18,19] the same with our results. In addition, although other studies showed an equal ratio of male and female in the adult population, [25,17] a female predominance was demonstrated in this large study.
In terms of treatment, gross totally surgical resection usually was not the primary treatment option in patients with ONGs, especially for low-grade at any age. [1,19,17]

Factors Associated with Survival and Tumor Management
Gender, marital status, race, year of diagnosis, tumor site, and different surgical patterns were not critical predictors of survival in univariate and multivariate analysis. However, age at diagnosis, tumor grade, and adjuvant RT and chemotherapy were signi cant survival factors in adult patients with ONGs. Although age was not recognized as a prognostic factor in a previous study, [19] our result showed that patients with age ≥ 50 years had a worse prognosis in both low-and high-grade groups, similar to other published series of gliomas. [29,30] It is universally acknowledged that patients with low-grade gliomas have a better survival prognosis, which had a 5-year OS rate of 85.5% compared to 10.5% for patients with highgrade gliomas in our study. Furthermore, although many reported cases of adult ONGs are high-grade, we found most of them remain low-grade tumors, inconsistent with the study by Shofty et al., in which 80% of primary adult ONGs are also low-grade. [17] Unlike the gliomas in other sites, there is no gold standard in treating both low-grade and high-grade ONGs. The general goal of all individualized treatment strategies is to preserve the patient's vision as long as possible. Based on this goal, surgical resection might not be a preferred option and is discouraged due to the inevitable blindness on the affected side or bilateral visual loss risks. [31] Our current results demonstrate no signi cant differences among various surgical options for patients with any grade ONGs regarding OS. Therefore, initial observation is frequently recommended with ophthalmological evaluation and neuroimaging surveillance to con rm clinical stability, particularly for low-grade ONGs. However, for malignant ONGs or progressive tumors with the aggressively visual decline, it might be appropriate to consider surgical excision to debulk the tumor and obtain a pathological diagnosis before commencing the following adjuvant therapy. [7,1,2] The role of adjuvant RT and chemotherapy in an adult population remains unclear due to its rarity. To our knowledge, there have been only nine documented cases of low-grade optic pathway gliomas in adults in the past few years. [17,16]  NF-1 status, or neuroendocrine morbidity, are unable to obtain from the SEER database. Finally, given the impossibly pathological and radiological review of the tumors, it is likely that some tumors may have been misdiagnosed owing to the low interobserver agreement in the diagnosis of different glioneuronal tumors. However, considering the prospective studies not available and not be expected in the near future, because of the scarcity of ONGs, a large retrospective study like this seems to be the best and most useful approach available to de ne the role of different adjuvant treatments for these lesions. Therefore, although several limitations mentioned above, this is the largest reported series and the best evidence available in regard to adult ONGs up to date.

Conclusion
This is the largest retrospective study of adult ONGs, including 179 cases from the SEER database. Our data rea rms that the low-grade ONGs have a signi cantly better prognosis than the high-grade, and demonstrates the different role of adjuvant treatments in different grade tumors. Although both adjuvant RT and chemotherapy have no positive effect on patients' OS in the high-grade group, they may have a potentially negative impact on patient's survival and prognosis in the low-grade group. Therefore, adjuvant RT or chemotherapy might not be recommended to consider in adult patients with low-grade ONGs, unless the malignant transformation or aggressive progression has been con rmed.

Declarations
Canada Clinical Trials G (2009) Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol 10 (5) Figure 1 Overall survival of adult patients with low-and high-grade optic nerve gliomas. Low-grade group. D-F. High-grade group.