Unravelling the gut-tumor axis in neuroblastoma: promising leads and cautionary gaps

Abstract

We read with great interest the article by Chu et al. investigating childhood neuroblastoma (NB) and gut microbiota (GM) using Mendelian randomization (MR) [1]. NB is the most common extracranial solid cancer among children [2], and novel insights into its etiology and causation are the need of the hour. The authors used genome-wide association study (GWAS) data from MiBioGen and IEU consortia to identify genus of GM which were associated with NB. They identified class Erysipelotrichia as potentially protective (IVW odds ratio ~ 0.37) and genus Oscillospira as a risk factor for NB (OR ~ 2.38). The authors also identified microbiome-linked host genes PELI2 and MUC4 which are associated with tumor pathways. It was interesting to find the proposition that short-chain fatty acids produced by Erysipelotrichia may promote neuronal differentiation in NB. This integration of clinical observations with mechanistic hypotheses represents an innovative direction in pediatric oncology research. However, several limitations merit discussion. A foremost issue is related to limited population diversity as the GWAS database was predominantly European. GM varies considerably with geographical location, ethnicity and diet patterns [3], hence causal inferences drawn in a particular population may not be applicable across diverse global populations. Future studies should incorporate more diverse cohorts to enhance generalizability of findings. An additional caveat is regarding the “snapshot” genomic data from the GWAS. There is no information regarding the tumor risk, stage, clinical characteristics and the modality of therapy received. It remains unclear as to whether GM plays a role in the initial tumorigenesis of NB or if the therapy received for the same alters the GM. NB patients are subjected to chemotherapy which is known to alter the GM [4]. In addition, use of antibiotics for management of complications arising out of chemotherapy can perturb the GM. It is difficult to draw inferences about chronology and causation without longitudinal genetic data. Another limitation lies in the resolution of the microbial data. The researchers studied the gut microbiome using 16S rRNA sequencing aggregated at the genus or class level. Such coarse taxonomic resolution may confound heterogeneity at the strain-level. For instance, “Oscillospira” is a poorly characterized genus-level bacterial taxon, and class “Erysipelotrichia” encompasses a taxonomically broad category. Additionally, 16S rRNA sequencing lacks detailed functional information pertaining to gene activity, metabolite production and host interaction. This precludes elucidation of the precise mechanistic pathways responsible for the observed associations. A more robust approach would have been the use of shotgun metagenomic sequencing [5] which has the ability to identify microbes down to the species or strain level. In addition, it enables a more detailed pathway level analysis. Utilization of metatranscriptomics and/or stool metabolomics in the present study would have enhanced the interpretability of the findings. Future studies could also utilize gnotobiotic animal models or integrative multi-omics for uncovering the precise microbiome-gene-pathway-host crosstalk, which might prove critical in our understanding of this complex topic, enabling validation of these associations. It could potentially pave the way for microbiome-modulating therapeutic strategies in NB.