The amplification of the gene MYCN (V-mycmyelocytomatosis viral-valeted oncogene, neuroblastoma derived) has been a well-documented indicator for poor prognosis in neuroblastoma, a childhood cancer. Unfortunately, there has been limited success in understanding MYCN functionality in the landscape of neuroblastoma and more importantly, given that MYCN has been deemed “undruggable,” the need to potentially illuminate key opportunities that indirectly target MYCN is of great interest. To this end, this work employs an emerging quantitative technique from network science, namely network curvature, to quantify the biological robustness of MYCN and its surrounding neighborhood. In particular, when amplified in Stage IV cancer, MYCN exhibits higher curvature (more robust) than those samples with under expressed MYCN levels. When examining the surrounding neighborhood, the above argument still holds for network curvature, but is lost when only analyzing differential expression – a common technique amongst oncologists and computational/molecular biologists. This finding points to the problem (and possible solution) of drug targeting in the context of complexity and indirect cell signaling affects that have often been obfuscated through traditional techniques.