Wilms tumor (WT), also known as nephroblastoma, is the second most common intra-abdominal tumor and the most common primary renal tumor in children (1). Approximately 75% of children with WT develop the condition between the ages of 1 and 5 years, most commonly at the age of (2). After years of clinical exploration, the five-year survival rate for WT has improved from less than 30% to 85%–90%. However, the recurrence rate remains at 15%–50% (3). The treatment and prognosis of nephroblastoma in children are related to histological staging (4), and five-year survival rates have reached 90% for children with good histological types of WT after moderate treatment (5, 6).

The treatment of nephroblastoma continues to be based on surgical resection supplemented by a combination of chemotherapy and radiotherapy (7, 8). However, treatment-related complications remain a problem for many children with WT, with treatment often triggering nausea and vomiting, loss of appetite, anemia, alopecia, and neutropenia, which subsequently affect patients’ psychological well-being (9). At the same time, surgical removal of the diseased kidney has limitations. Therefore, current treatments are not entirely appropriate for some populations, especially infants and children and patients with bilateral tumors (10). Therefore, the key to improving patient prognosis is to improve treatment based on clinical and biological risk factors, and further stratification of current treatment options based on the prognostic value of tumor biology would be an important approach to improving WT prognosis (11, 12). Along with the development of CRISPR/Cas9 gene-editing technology, artificially modified chimeric antigen receptor T cell immunotherapy, and aptamer technology, precise genetic and biological therapies could be a new option for the treatment of nephroblastoma (13–15).

With the development of second-generation sequencing, the discovery of new tumor markers and potential therapeutic targets has become more accessible. Advances in RNA sequencing technologies have revealed the complexity of the human genome. The study of the RNA transcriptome is one of the most important challenges facing biology today, as RNAs represent new potential biomarkers and drug targets (16, 17). Currently, a growing number of studies on WT have identified many key mRNAs that are closely associated with the prognosis of this tumor (18, 19). It is well known that inter-individual heterogeneity usually constitutes only a more traditional prognostic system. For example, risk stratification based on the TNM staging system alone is not sufficient, nor is it sufficient to provide an accurate prediction of survival outcomes. A study by Lin et al. (20) identified a 5 mRNA signature as a new potential prognostic biomarker for WT, beyond which models have not been over reported. Therefore, additional prognostic models are needed to predict survival outcomes in pediatric WT patients.

This study was a comprehensive study to analyze differential genes in WT samples through multiple datasets and to develop validated gene signatures for predicting prognosis in WT patients, as well as to further screen key genes to provide a research basis for future biological treatment and clinical diagnosis of WT.

Numerous studies have reported that mRNA plays a crucial role in the tumorigenesis development of WT (24). However, with the development of detection techniques, a single mRNA expression pattern is no longer sufficient to accurately predict the prognosis of WT. In addition, the role of miRNAs in altered gene expression should not be neglected. Zhu et al. (25) reported that miR-92a-3p inhibited the proliferation, migration, and invasion of WT cells by regulating the NOTCH1 signaling pathway. Therefore, the identification of differentially expressed miRNAs represents a promising strategy. However, heterogeneous results are primarily generated by studies with relatively limited sample sizes, several candidate miRNAs, or a lack of experimental validation. In this study, we identified one hub miRNA—miR-30c-5p—by integrating the dataset and constructed a 12-gene marker-based prognostic prediction model for WT based on the prediction of miR-30c-5p target genes and the results of WT differentially expressed gene analysis.

MiR-30 c-5p (Previous ID: miR-30c) was first identified in 2002 by Lagos-Quintana et al. (26) from mouse heart and brain tissues, and its sequence is highly conserved across species. Several studies have reported that miR-30c-5p is aberrantly expressed in different tumor tissues, sera, and cell lines and is associated with clinical features and prognostic factors in a variety of cancers, including lung cancer (27), breast cancer (28), and colorectal cancer (29). In 2010, Heinzelmann et al. (30) screened 12 miRNAs, including miR-30c-5p, which were lowly expressed in highly invasive renal clear cell carcinoma with early metastasis. In an in-depth study, it was found that miR-30c-5p expression levels were significantly lower in primary tumors with metastasis compared to normal kidney tissue and primary renal clear cell carcinoma without metastasis, and miR-30c-5p expression levels were significantly correlated with the 5-year progression-free survival of patients (31). Thus, miR-30c-5p may be a useful indicator for the early prediction of renal cancer metastasis, and different expression levels may be associated with specific distant metastasis. This study focused on the role of the target genes of miR-30c-5p in WT. We predicted the target genes of miR-30c-5p and obtained a total of 384 genes in combination with differentially expressed genes in TARGET-WT. Furthermore, KM survival analysis of these 384 genes was performed to obtain 25 genes associated with cervical cancer prognosis, and LASSO Cox regression was used as a machine-learning algorithm to construct a prognostic risk model with 12 gene signatures (i.e., BCL6, CCNA1, CTHRC1, DGKD, EPB41L4B, ERRFI1, LRRC40, NCEH1, NEBL, PDSS1, ROR1, and RTKN2), and the patients were divided into high-risk and low-risk groups according to risk scores. The prognosis of patients in the high-risk group was significantly worse than that in the low-risk group, and the working curves of subjects were used to verify that the model had good predictive ability for 1, 3, and 5-year prognoses of WT patients, and the risk group could be used as an independent predictor of WT prognosis in patients. Previously, gene and miRNA signatures were developed for WT prognosis prediction. However, the number of models is far from enough.

Finally, we performed univariate and multivariate Cox analyses for each of the 12 genes in the model and found that ERRFI1 and ROR1 served as independent predictors. ERBB receptor Feedback Inhibitor 1 (ERRFI1), the product of mitogen-inducible gene 6, through its ERBB-binding region, docks with the EGFR kinase structural domain docking, inhibiting EGFR activation and downstream signaling (32). We found low expression of ERRFI1 in WT cancer tissues after comparing its expression in WT, and it has been reported that ERRFI1 is also frequently mutated or downregulated in breast cancer (33), lung cancer (34), and glioblastoma (35). However, we found that high ERRFI1 expression was significantly associated with a poor prognosis in WT patients.

ERRFI1 is considered a tumor suppressor that directly inhibits epidermal growth factor receptors. However, new studies have found that the role of ERRFI1 depends on EGFR levels; therefore, in a low EGFR setting, downregulation of ERRFI1 leads to higher migration rates and promotes cell growth (36). In the Jäger K et al. (37) human study, ERRFI1 upregulation was found to be significantly associated with poor prognosis in metastatic melanoma, and in the present study, ERRFI1 upregulation was similarly found to be associated with poor prognosis in WT. Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is a member of the ROR family of type I transmembrane receptors with ligand-bound extracellular and intracellular tyrosine kinase domains.

During embryogenesis, ROR1 plays a physiological role in neural, auditory, skeletal, and vascular organogenesis, but studies have shown that ROR1 is absent or expressed at low levels in most adult tissues (38, 39). However, as an oncoprotein, ROR1 can reemerge in hematological and solid tumors, especially in histologically advanced tumors, where ROR1 may promote tumor cell migration through Wnt5a signaling or interaction with other receptors (40, 41). In the present study, we found high expression of ROR1 in WT cancer tissues. Hodjattallah Rabbani et al. (42) verified that detecting a high level of ROR1 expression in blood cells may help in the early detection of renal malignancies. Notably, Kaplan–Meier survival analysis found that low ROR1 expression was significantly associated with poor patient prognosis, a result that is contrary to most studies that have partially demonstrated poor prognosis in patients with high ROR1 expression, such as ovarian cancer (43), colorectal cancer (44), and so on. Therefore, the prognostic predictive role of ROR1 expression in WT remains to be explored in depth.

In summary, this study constructed a 12-gene signature prognostic risk model based on the target genes of miR-30c-5p and determined that ERRFI1 and ROR1 could be used as independent predictors of WT prognosis in patients. However, this study has several shortcomings. These include the lack of biological behavior studies for the identified ERRFI1 and ROR1 and the fact that there are few studies on ERRFI1 and ROR1 in WT. However, it also has potential research value. In addition, WT is a relatively rare primary malignancy, and more clinical samples and survival information are needed to validate the results of this study. Although the risk model constructed in this study displayed better performance, the LASSO Cox algorithm used will adjust the parameters appropriately during the calculation to avoid degradation of the algorithm’s performance (45). Therefore, more accurate machine-learning methods, a larger number of clinical samples, and in vitro experiments need to be further developed in future studies.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.