Purpose or Case Report: Radiogenomics refers to the correlation of imaging and genomic data of tumors in cancer patients. This study attempts to make correlations between biomarkers in NGS Comprehensive Solid Tumor Panel reports and MRI imaging findings in neuroblastoma patients.
Methods & Materials: Patients aged 0-7 yrs who had NGS Solid Tumor Panel testing and desired imaging sequences were compiled into a database (N=25). Primary tumor data from axial fluid sensitive sequences; T2 fse fat sat (23), Haste (1), and STIR (1) from PACS were processed using the tumor segmentation feature in the pMRI software package. Tumors were segmented by hand and segmentation was guided/approved by a radiologist with 20+ years of experience. Volume and signal intensity values were collected using pMRI. Signal intensity/heterogeneity is defined by standard deviation, kurtosis, or skew. Genetic data were collected from Solid Tumor Panel reports in EPIC.
Results: 48% (n=12) of patients were male and 52% (n=13) were female. Tumor volumes (V) ranged from 2.85-521.97 cm³. Mean age was 2.6 yrs. Independent samples T tests and two-way ANOVAs were run to correlate statistically significant findings.
Analyses showed 2p addition tumors were larger than those without: Yes; n=7, volume=333.86 cm³. No; n=18, volume=83.42 cm³, p=.009. 11q deletion tumors were larger than those without: Yes; n=5, V=305.26 cm³. No; n=20, V=115.625 cm³, p=.017.
11q deletion + 2p addition (n=3) tumors were larger than 2p addition but no 11q deletion tumors (n=4). ηp²=.29. F(1, 21)=8.792, p=.007.
MYCN amplification indicated larger tumors: Yes; n=6, V=307.5 cm³. No; n=19, V=104.9 cm³, p=.046. 1p deletions indicated larger tumors: Yes; n=6, V=203.005 cm³. No; n=19, V=103.67 cm³, p=.001.
1p deletion but no MYCN amplification (n=1) indicated larger tumors than having neither mutation (n=18). MYCN amplification but no 1p deletion (n=5) indicated larger tumors than having neither (n=18). ηp²=.25. F(1,21)=6.99, p=.015.
Heterogeneity, kurtosis, and skew of signal intensity did not meet significance.
Conclusions: Tumor volume is significant in poor prognosis genes. Collecting/analyzing these data provides detailed tumor profiles and useful information for targeted therapy development. Correlating imaging findings and genetics provides less invasive analyses and prognosis indicators. Other trends were noticed but did not meet significance. With a larger dataset, these could be proven and clinically applied. This is a future goal of this study.