Poor Treatment Outcome of Neuroblastoma and Other Peripheral Nerve Cell Tumors May be Related to Under Usage of Radiotherapy and Socio-Economic Disparity : A US SEER Data Analysis

The Surveillance Epidemiology and End Results (SEER) cancer registry has been extensively used to modeling outcome prediction models peripheral nerve cell tumors (PNCT) including neuroblastoma (NB). Numerous studies have done to better characterize these rare tumors to identify socio-economic disparity in treatment outcome and to build models for selecting patients for clinical trials (Esiashvili et al., 2007; Benoit et al., 2008; Hsieh et al., 2009; Shapiro and Bhattacharyya, 2009; Friedman et al., 2010; Pan et al., 2010: 2011; Bhatia, 2011; Johnson et al., 2011; Navalkele et al., 2011; Platek et al., 2011; Pui et al.,


Introduction
The Surveillance Epidemiology and End Results (SEER) cancer registry has been extensively used to modeling outcome prediction models peripheral nerve cell tumors (PNCT) including neuroblastoma (NB).Numerous studies have done to better characterize these rare tumors to identify socio-economic disparity in treatment outcome and to build models for selecting patients for clinical trials (Esiashvili et al., 2007;Benoit et al., 2008;Hsieh et al Surveillance Epidemiology and End Results (SEER) (http://seer.cancer.gov/) is a public use cancer registry of United States of America (US).SEER is funded by National Cancer Institute and Center for Disease Control to cover 28% of all oncology cases in US.SEER started collecting data in 1973 for 7 states and cosmopolitan registries.Its main purpose is through collecting and distributing data on cancer, it strives to decrease the burden of cancer.SEER data are used widely as a bench-mark data source for studying PNCT outcomes in US and in other countries (Perme and Jereb, 2009;Lacour et al., 2010;Johnson et al., 2011).The extensive ground coverage by the SEER data is ideal for identifying the disparity in oncology outcome and treatment in different geographical and cultural areas for cancers (Bhatia, 2011;Johnson et al., 2011;Pui et al., 2012).In addition to the biological staging factors and the treatment factors, this database also contains a large number of county level socio-economic factors data.This study aimed to identify barriers to good treatment outcome that may be discernable from a national database.

Materials and Methods
SEER registry has massive amount of data available for analysis, however, manipulating this data pipeline could be challenging.SEER Clinical Outcome Prediction Expert (SCOPE) (Cheung, 2012) was used mine SEER (Cheung et al., 2001a, b).The data were obtained from SEER 18 database.SEER is a public use database that can be used for analysis with no internal review board approval needed.SEER*Stat (http://seer.cancer.gov/was: Site and Morphology.ICCC site recode ICD-O-3} ='IV Neuroblastoma and other peripheral nervous cell tumors'.This study explored a long list of socio-economic, staging and treatment factors that were available in the SEER database. The codes of SCOPE are posted on Matlab Central (www.mathworks.com).SCOPE has a number of utility programs that are adapted to handle the large SEER data pipeline.All statistics and programming were performed in Matlab (www.mathworks.com).Each risk factor outcome (cause of death: other neuroendocrine including thymus as coded in SEER).The areas under the receiver operating characteristic curve (ROC) were computed.
if the ROC performance did not degrade (Cheung et al., 2001a, b).In addition, it also implemented binary fusion and optimization to streamline the risk stratification by combining risk strata when possible.SCOPE uses Monte Carlo sampling and replacement to estimate the modeling errors and allows t-testing of the areas under the ROC.SCOPE provides SEER-adapted programs for user friendly exploratory studies, univariate recoding and parsing.

Results
There were 5261 patients included in this study (Table 1).There were Neuroblastoma (n=3742, 71.11%),Other pediatric and embryonal tumors, NOS (n=743, 14.2%), and Paraganglioma and glomus tumors (n=776, 14.76%).including thymus'.The follow up (S.D.) was 83.8 (97.6) months.47% of the patients were female.The mean (S.D.) age was 18.04 (25) years.Children and young adults constituted two third of the PNCT patients listed from SEER data.About 34.45% of patients were grouped in the un-staged/other categories.For pretreatment factors, the SEER staging categories using localized, regional, metastatic and un-staged (Table 1) has the highest ROC (S.D.) area of 0.58 (0.01) among the factors tested in Table 1.Expert (SCOPE) was used to perform ROC curve and area under the curve calculations (Figure 1).In this example, the ROC area of the 4-tiered SEER staging model as computed for 5 random samples (Figure 1 upper panels the 4-layered SEER risk levels (local, regional, distant, un-staged) to a simpler non-metastatic (I and II) versus metastatic (III) and un-staged (IV) model.The ROC area (S.D.) of the 3-tiered model was 0.59 (0.01) based on 5 random samples with replacement from the SEER data.Whether the patient received surgical treatment was the most predictive factor among treatment factors and overall metropolitan residence was 20.8% versus 25.9% (Table 1 and 2).African American PNCT patients had 36% risk of ethnicity groups Tables 1 and 2).The level of differences for these two factors did have make the socio-economic factors very predictive of outcome.They had a ROC area of around 0.5 that is expected for a random variable with no predictive power.County's family income level and county's education attainment did not contribute to poor outcome.

Survival of PNCT (including NB) Patients
70% patients did not receive RT (Table 1).Figure 2 shows that even in the localized and regional stages when RT could be used for curative intent (Platek et al., 2011), very low percentage of patients under went RT. Fig. 3 shows that this under usage of RT is most evident among the adult patients when the radiation side effects  are expected to be less severe compared with younger patients (Miralbell et al., 2002;Cohen et al., 2005;Herzog, 2005).

Discussion
This study is interested in constructing models that will aid patient and treatment selection for PNCT cancer patients.To that end, this study examined the ROC models (Hanley and McNeil, 1982) of a long list of potential explanatory factors (Table 1).ROC models take into Ideal model would have a ROC area of 1 and a random model is expected to have an area of 0.5 (Hanley and McNeil, 1982).For example, a clinical ROC model can be used to predict if a patient receiving the recommended treatment will die from the disease.The SEER anatomic staging is most predictive pretreatment model of patient outcome (Figure 1 and Table 1).Therefore it is useful in guiding treatment selection.After binary fusion, it reduced to non-metastatic versus metastatic versus un-staged for clinical trials because it has much fewer risk groups than the current PNCT grouping to balance (Platek et al., 2011;Pui et al., 2012).
When there are competing prediction or prognostic  thought to prevail (D'Amico et al., 1998).This has an information theoretic under-pinning.For practical purposes, simpler models require fewer patients for a randomized trials because fewer risk strata need to be balanced.In the clinic, simpler models are easier to use.SCOPE streamlined ROC models by binary fusion (Table 1).Two adjacent strata were tested iteratively to see if predictive power usually belong to the more complex models.This study has shown that SCOPE can built For radiotherapy, the ROC areas were modest (0.5).Low ROC areas imply the information content (i.e. the staging accuracy) of the models may be limited.It is consistent with the fact that only 70% patients had complete SEER staging (Table 2).In addition, the outcome of the completely staged patients was much more superior when compared with the entire cohort (Figure 2).It may be a consequence of having a better guidance model in treatment and patient selection.PNCT cancers have good treatment outcomes, there is a 4-10% risk of PNCT death (Table 2) at a localized/regional stage, however, the under-staged patients were disadvantaged and had double this risk likely due to the lack of guidance in treatment selection.There was only 15%-35% use of RT in the localized and regional PNCT patients.These data suggest an under treatment of the adult PNCT patients (Figure 3) with RT.And outcome could be further improved in these patients, especially when they are adult patients.Thus radiation oncologists should be more attentive in recommending RT for these patients.For the pediatric populations, proton use is expected to improve the outcome of these patients by primarily decreasing the rate of secondary cancers (Bassal et al., 2006;Schultz et al., 2007;Friedman et al., 2010;Kuhlthau et al., 2012).models are the most prognostic pre-treatment model of treatment outcomes of PNCT cancer patients.The high under-staging rates may have prevented patients of radiotherapy after surgery use may have contributed to the poor outcome in these patients with this disease.Furthermore, education and access to appropriate cancer treatment may eliminate the potential socio-economic barriers to good outcome.
PCNT models to aid treatment and patient selection.

Figure 2 .
Figure 2. Fraction of Patients Received Radiotherapy at each SEER Stage