Primary optic nerve glioma, or optic pathway glioma (OPG), is the most common primary tumor of the optic nerve (1) and arises from astrocytes, a glial cell subtype found in the central nervous system (CNS) (2,3). The majority of these neoplasms present as low-grade tumors and are mostly seen in the first decade of life (1). Additionally, OPGs are strongly associated with neurofibromatosis 1 (NF1); OPGs are observed as the most common CNS tumor within this disease entity (4,5). The clinical rationale for patients diagnosed with primary OPG includes genetic screening for NF1. Diagnosis is confirmed through imaging, preferably magnetic resonance imaging (6), albeit this neoplasm may be evident through a computed tomography scan (7). Importantly, biopsy is unwarranted due to inherent and enhanced imaging sensitivity for typical pathognomonic findings of OPGs on imaging (8). The clinical course of OPGs can be volatile in children and adults due to their location and variance in histology, respectively (9-13). Currently, controversy exists regarding what the most optimal treatment regimen for OPG may be (14,15).

The occurrence of SPMs is common as they are diagnosed in one out of every 12 patients with cancer (16). A multitude of prior studies have characterized the risk of SPM development in the setting of a variety of primary cancer, including primary sites of the breast, colon, and lung (17-19). Furthermore, it was noted that more than a half of patients diagnosed with two types of cancer died due to their SPM (20). It is reported that the 5-year survival rate associated with OPG is more than 90% (21,22). While the survival rate is impressive, clinical consequences include a heightened propensity for development of a SPM (23-25). One study showed a significantly increased rate of SPM in children diagnosed with NF1 and treated for OPG with radiotherapy (26).

The number of patients diagnosed and treated for OPGs is increasing. Despite continual improvements in survival, it is important to understand the long-term complications of malignancy. To our knowledge, few studies characterize the incidence of SPM in those with first primary OPG. This retrospective cohort study aimed to analyze the occurrence and risk of SPM in patients with a diagnosis of OPG between 1992 and 2017 through The Surveillance, Epidemiology, and End Results (SEER) database.

Within the SEER system, the “Incidence-SEER Research Data, 13 Registries (excl AK), Nov 2019 Sub (1992-2017)” database was analyzed for patients diagnosed with optic nerve gliomas between January 1992 and December 2017. This retrospective cohort study filtered over 10 million patients by utilizing the search criteria “C72.3-Optic Nerve” for “Site and Morphology, Primary Site-labeled) and selecting only cancer with malignant behavior. Multiple outcome analysis was performed yielding standardized incidence ratios (SIR) for SPM, and excess absolute risk (EAR). The 95% confidence intervals for SIRs are given with statistical significance achieved at p<0.05.

Cohort analysis. Descriptive statistics of a total of 622 patients with a formal diagnosis of primary glioma originating within the optic nerve are given in Table I. Patients had a mean age ±SD of 62.20±12.52 years (range=0.5-80). Please see Table I for additional information regarding baseline demographical characteristics of this cohort.

For the majority of patients, diagnostic confirmation was obtained via radiography without microscopic confirmation (63.2%), while positive histology (32.3%), direct visualization without microscopic confirmation (1.1%), and clinical diagnosis (1.8%) were less often utilized. Nearly all cases (97.7%) were reported from hospital inpatient/outpatient services or clinics. Within the subdivision of patients acquiring a SPM, the mean age at diagnosis of primary OPG was 13.52 years of age and of SPM was 26.66 years of age.

SIR of SPM. Relative to the US national population, patients afflicted with primary OPG demonstrated a significantly increased risk for multiple SPMs. Primary malignancies originating from soft tissues (including the heart) (SIR=33.23, CI=6.85-97.11; EAR=5.07), breast (SIR=4.99, CI=1.36-12.77; EAR=5.57), brain (SIR=105.38, CI=65.23-161.08; EAR=36.23), cranial nerves (SIR=103.29, CI=12.51-373.12; EAR=3.45), non-lymphocytic leukemia (SIR=15.05, CI=1.82-54.37; EAR=3.25), myeloid and monocytic leukemia (SIR=16.26, CI=1.97-58.75; EAR=3.27), and Kaposi’s sarcoma (SIR=79.88, CI=2.02-445.08; EAR=1.72) demonstrated significantly increased SIRs as SPMs (Table II). In totality, cumulative SPM (SIR=6.04, CI=4.33-8.19; EAR=59.60) highlighted overall significance in incidence rates of SPM in patients diagnosed with OPGs.

Utilizing a large-scale population-based cancer database, this study evaluated age-adjusted SIRs of SPMs in patients with a first primary diagnosis of OPG. Analysis of baseline demographic data showed 61.7% of our population was diagnosed with primary OPG before 10 years of age, corroborating previous studies highlighting 70% of OPGs are diagnosed in the first decade of life (1,16). The overall risk of SPMs for all sites was significantly elevated in patients with primary OPGs (SIR=6.0; 95% CI=4.3-8.2) (Table II). Risk was also increased for all types of solid tumor, soft-tissue tumors (including heart), breast tumors (when considering both male and female patients, and females alone). Interestingly, brain and cranial nerve tumors had the highest SIRs (SIR=105.4, CI=66.7-157.9; and SIR=103.3, 95% CI=12.5-373.1, respectively) (Table II).

Previous studies have been limited in exploring the relationships between these primary malignancies (27). However, hypotheses describe shared embryonic etiologies to be the greatest contributing factor to the reported increase in brain and cranial nerve tumors as has been reported for various other cancer types (27-29). A multitude of mutual risk factors found in cancer of the neuroepithelium (e.g., ionizing radiation) might also contribute to the increased age-adjusted SIR for brain and cranial nerve tumors (27,30,31). Lastly, genetic mechanisms influence the association between OPGs and SPMs of the brain and cranial nerves (32). Multiple genetic syndromes (e.g., NF1 and -2), specific genotypes (i.e., glioma CpG island methylator phenotype) and certain molecular mutations (BRAF proto-oncogene - V600E mutation) are well-characterized and implicated in both OPGs and brain/cranial nerve malignancies (32-34).

Unexpectedly, our survival study discovered a high incidence of Kaposi’s sarcoma in patients with primary OPGs (SIR=79.88, CI=2.02-445.08), not previously reported as far as we are aware. One study showed that human herpesvirus-8 (HHV-8) infection modulates the proliferation of glioma stem-like cells (GSCs) in vitro (35). After establishing both HHV-8-infected glioblastoma cell and GSC lines, HHV-8 was associated with significantly higher infectivity for GSCs compared to human endothelial cells, and higher rates of cellular proliferation and lower cell death in GSCs. One proposed pathway for this is HHV-8 microRNA-accelerated phosphorylation of AKT (36).

Non-lymphocytic leukemia (SIR=15.1, CI=1.8-54.4; EAR=3.3) and myeloid and monocytic leukemia (SIR=16.3, CI=1.97-58.8; EAR=3.3) also demonstrated significantly increased incidence in our study population (Table II). Some studies have shown a strong involvement of pre-B-cell leukemia homeobox 3 (PBX3) transcription factor in gliomas and in multiple other cancer types, especially leukemia (37-40). A 2017 study elucidated how silencing PBX3 reduces proliferation of glioma cells in vitro and in vivo (38). Furthermore, how microRNA-98 attenuation of PBX3 reduces cell migration and invasion in glioma was also outlined (38). PBX3 overexpression has frequently been reported in many solid tumor types and in several hematological malignancies, serving as a promoter for cell survival, proliferation, and invasion (39). Further gene and protein expression data are needed to fully understand the relationship between optic nerve gliomas, PBX3 and the development of SPMs.

The limitations of our study are several and merit understanding before implications of the results can be considered. Firstly, the rate of SPMs may have been inflated given that metastases from the primary glioma might have been misclassified as second malignancies. Additionally, sites of increased secondary malignancies (soft tissue including of the heart, breast, and Kaposi’s sarcoma) are unlikely locations for distant metastasis of OPGs which tend not to metastasize (41). While rigorous quality control and data gathering protocols by the National Cancer Institute allow for reasonable generalizability, this comprehensive database only accounts for roughly 28% of the entire US general population. Consequently, rare SPMs of OPG such as Kaposi’s sarcoma, are possibly unaccounted for. Lastly, due to inherent limitations of the SEER database, adjustments for common cancer risk factors such as consumption of tobacco or alcohol, the presence of obesity, or of underlying genetic mutations were not made. Such unaccounted characteristics are undoubtedly associated with both primary and secondary cancer and would certainly modify the incidence of given SPMs.

This study highlights significantly increased SIRs for SPMs in patients diagnosed with OPGs in a nationally representative database via a multiple outcomes analysis. These primary cancer types include originations from the breast, CNS, cranial nerves, non-lymphocytic leukemia, myeloid and monocytic leukemias, and Kaposi’s sarcoma. Increased sensitivity in clinical diagnosis coupled with modern advancements in cancer treatment confer early detection and increased survival, leading to greater incidence of SPMs (30,31,42,43). Clinical surveillance utilizing real-time screening measures in this predisposed target population can promote longevity and health in patients diagnosed with OPGs.

The Authors declare that they have no competing interests.

Z.H. performed statistical analysis of SPMs in patients with primary optic nerve gliomas. J.K. and A.S. contributed equally to the drafting of the article. F.D. provided direction via clinical oversight and mentorship during project completion. All Authors read and approved the final article.