DISCUSSION
Several somatic genetic alterations, that provide critical prognostic
information, have been described in NB. Among those, MYCN amplification
is a crucial prognostic factor, being associated to advanced tumor stage
and poor outcome, and is central to the risk stratification systems.
According to the recommendations of the INRG Biology Committee, the
methods accepted for detecting MYCN amplification are represented by
FISH, Polymerase Chain Reaction (PCR), Array comparative genomic
hybridization (aCGH) and Multiple Ligation dependent Probe Amplification
(MLPA) on tumor tissue [17]. In clinical practice, the availability
of tumor material is sometimes challenging even if biopsy remains the
gold standard for the diagnosis of NB. Several minimally invasive
surgical approaches have been adopted, including samples from
fine-needle aspirates, but they result to be burdened by a significant
risk of hetMNA, giving clinicians only partial information that could be
misleading in the choice of the most appropriate therapeutic approach.
Furthermore, MYCN amplification may rarely show hetMNA itself, making
the sampling and following diagnosis even more challenging. Recently,
other experimental methods to detect MYCN amplification by using
cell-free DNA in serum and plasma have been developed in patient
affected by NB [18, 19]. However, these approaches have not been
validated yet in the current clinical practice. In this context,
radiomics and radiogenomics, may play a significant role thanks to the
quantitative noninvasive and repeatable analysis of standard clinical
imaging that encompasses the whole tumor volume, taking into account the
hetMNA. Radiomics features may be associated with specific molecular
pathways or mutations, offering reliable predictive tools for clinicians
in breast cancer [20], glioma [14], lung cancer [15, 21] and
even in healthy tissue [22]. Furthermore, CT represent the most
common modality used to diagnose and stage NB and for this reason the
images are easily accessible even for retrospective analysis. Brisse and
colleagues correlate semantic features to the genomic profile of NB,
showing a significant relationships among the sympathetic origin of the
tumor, its genomic profile and the outcome [23]. In our preliminary
study we aimed to develop a radiomics single phase CT based predictive
model as a convenient and reliable biomarker for MYCN amplification
status in NB. The proposed radiomics signature was set up to
differentiate among MYCN amplified and MYCN wild types tumors,
representing a promising tool for prognostic prediction in a
particularly innovative scenario. Wu and colleagues have recently
published a similar model, achieving similar results in MYCN prediction
using a combined clinical-radiomics modeling approach. Despite the
similar structure of the two experiences, our model focuses only on one
CT sequence, reducing the foreseen workload for a real-world application
of our approach in clinical practice and is based on a lower number of
radiomics features (2 versus 7), reducing the computational burden and
offering a more immediate workflow [24].
In conclusion, this preliminary study suggests that standard
post-contrast CT logistic radiogenomics classifier may help to identify
MYCN amplification status, which may successfully integrate the
traditional invasive MYCN status testing and provide survival
predictions for patients presenting amplification. The limitations of
this study are represented by the relatively small sample size, which
cannot reflect the overall MYCN status population, and the retrospective
and single-institution nature. Further prospective multicenter
validation studies are necessary to confirm the generalizability of the
observed results and to externally validate the proposed model.