Discussion
As few studies have established nomogram for predicting the survival of
patients with brainstem glioma patients, the sample size involved was
small, and the prognostic factors have been limited13.
Thus, we developed a clinical nomogram to predict the survival based on
SEER database. The SEER registry is the largest population-based
database of cancer patients in the United States, covering approximately
26% of patients diagnosed with cancer in the nation. We reviewed
patients’ data from the latest version of the SEER as released in 2015
(covering 18 registries, 1973-2015), by using SEER*Stat version 8.3.5,
and we also set a strict inclusion and exclusion criteria.
In this study, patient survival was predicted using a clinical nomogram
that was based on SEER database. This was necessitated because the
traditional staging classification, which is commonly used for survival
predicting and clinical strategies selecting for patients with cancers
cannot accurately and consistently distinguish the difference in
survival among various stages. The nomogram is a comprehensive,
accurate, and useful prognostic model, which has been previously used
for many kinds of malignances. However, in those studies, the
limitations were small sample sizes and analysis of few prognostic
factors. Five independent prognostic factors: age, race, tumor size,
grade and radiotherapy, identified through univariable and multivariable
Cox-Regression analyses were incorporated in the clinical nomogram.
Tumor histology grade category contributed significantly to prognosis,
according to the nomogram, which is consistent, but not identical, with
previous studies on survival risk factors for glioma patients, where
poorly differentiated and undifferentiated status were highly associated
with poor prognosis in children stem glioma patients. Most people with
low-grade gliomas are treated with surgery and may receive radiotherapy
thereafter. However, in this study, we found that pediatric brainstem
gliomas patients had no survival benefits from radiotherapy. We also
found that chemotherapy, was not an effective measure to improve outcome
in cases retrieved from the SEER database14,15. There
was a relatively significant effect of race on patient survival with
longer median survival times in white people compared to black people.
For validation of the nomogram to guarantee that the model could be
generally applied and to avoid overfitting, it was necessary to evaluate
discrimination using the C-index and calibration, which was assessed by
comparing the agreement between predicted and actual survival of
patients16,17. Our nomogram discriminated and
predicted survival more efficiently than the traditional staging system.
Further, the decision curves analysis showed that our model had a better
clinical net benefit across all threshold
probabilities18,19. Moreover, the risk stratification
system applied to two risk groups of patients could discriminate CSS in
children stem glioma.
The strength of our analysis of the risk stratification system is that
the nomogram was an accurate and reliable prognostic model that could
aid clinicians identify high-risk patients for targeted adjuvant
treatment, particularly for our highly selective
cohort20. There were some certain limitations in our
study. First, although we performed multivariable analysis to minimize
confounder effect associated with the heterogeneities, this was a
retrospective analysis, which was further compromised by the small
sample size and must be accounted for when interpreting the
results21. Second, the retrospective analysis may have
introduced the possibility of selection bias in the study
design22. Third, the SEER database lacks information
on modern gene-array technology and molecular biomarkers, such as status
IDH1/TERT expression23-26, which have proven to be
associated with CSS in children stem glioma patients. Therefore, future
prospective analysis is warranted to predict survival of pediatric cases
of gliomas.