Characteristics of patients

Following screening, a total of 184 patients with WT were included in this study and the base line data were shown in Table 1. 184 Patients were found to have 162 (88.04%) favorable histology (FH), 99 (53.80%) females and 85 (46.20%) males, 88 (47.83%) left and 96 (52.17%) right. There were 33 patients with adverse events: 10 (5.43%) recurrence, 19 (10.33%) death, and 14 (7.61%) metastases which include 9 (4.89%) pulmonary, 1 (0.54%) hepatic, 1 (0.54%) intestinal, 1 (0.54%) lung plus thoracic with mediastinal, 1 (0.54%) pelvic plus hepatic and 1 (0.54%) thoracic plus abdominal metastases).

Table 1 Characteristics of patients with WTAnalysis of prognostic factors in WTPeripheral blood cells compared between patients with WT and non-WT children

To investigate potential differences in blood cells of the children with WT and non-WT, we carried out the Wilcoxon rank test. The results showed that there were no significant differences in age and gender between the two groups (P > 0.05). However, there were notable variations in various blood cell parameters. WBC, PLT, N and N% in WT patients were higher than non-WT individuals, while HB, RBC, LB, LB%, M, M% were lower (P < 0.05). (As in Fig. 2a, we only show the key information in the main text, and detailed information can be found in Table S1 of the additional file.)

Fig. 2

The Wilcoxon rank test and survival analysis. a Histogram of the analysis of variance for WT and non-WT children; b The Kaplan–Meier curve for overall survival rate in WT patients; c The Kaplan–Meier curve for event-free survival rate in WT patients; d Forest graphs of OS and EFS in WT patients. * P < 0.5, ** P < 0.01, *** P < 0.001, WBC white blood cell count, PLT platelet count, RBC red blood cell count, HB hemoglobin, LB absolute lymphocyte count, N absolute neutrophil count, M absolute monocyte count, LB% lymphocyte percentage, N% neutrophil percentage, M% monocyte percentage, PLR platelet-lymphocyte ratio, NLR neutrophil–lymphocyte ratio, P P-value, n number, OS overall survival, EFS event-free survival

Analysis of the correlation between peripheral blood cells and tumor survival outcomes

Log-rank test was performed to compare blood indexes and survival outcomes (death or survival, event-free and positive events). As shown in Table 2, Monocyte count (M) below 0.325 × 103/μL (P = 0.029) and NLR above 1.380 (P = 0.029) accounted for more of the patients who died. Patients who had positive events presented a higher percentage of PLR above 94.632 (P = 0.035), LB below 3.570 × 103/μL (P = 0.029) and M below 0.325 (P = 0.035) × 103/μL than those with event free.

Table 2 Correlation analysis of pretreatment peripheral blood cells and tumor survival outcomesThe Kaplan–Meier survival curves of peripheral blood cells before treatment

The Kaplan–Meier survival curves for blood cells are displayed in Fig. 2b-c. (In the main text, we present only significant results; additional information is available in Fig. S1-S2 in the additional file.) In the entire study population, there were 19 (10.33%) deaths, with a 5-year OS of 90.22% and a 5-year EFS of 83.15%. Factors demonstrating significantly inferior overall survival by the log-rank test were NLR above 1.380 (HR 0.37, HR95% CI 0.15 ~ 0.91, P = 0.030), stage IV (P < 0.001) and M below 0.325 × 103/μL (HR 2.63, HR95% CI 1.07 ~ 6.47, P = 0.035). Simultaneously, higher EFS was associated with stage IV (P < 0.001), Age ≤ 3 years (HR 0.36, HR95% CI 0.18 ~ 0.73, P = 0.005), PLR below 94.632 (HR 0.47, HR95% CI 0.24 ~ 0.93, P = 0.030), LB above 3.570 × 103/μL (HR 2.24, HR95% CI 1.13 ~ 4.45, P = 0.021) and M above 0.325 × 103/μL (HR 2.22, HR95% CI 1.11 ~ 4.44, P = 0.024).

Multivariate cox regression analysis of prognostic factors correlated with WT

Variables with P < 0.05 in log-rank test were included in the multifactor COX analysis, and forest plots were developed (see Fig. 2d. The corresponding tables are available in Table S2-S3 in the additional file). A total of 2 factors were independently correlated with OS and EFS: M (HR 0.22, HR95%CI 0.08 ~ 0.62, P = 0.004 and HR 0.44, HR95%CI 0.20 ~ 0.95, P = 0.036, respectively) and stage IV (HR 7.89, HR95% CI 1.65 ~ 37.77, P = 0.010 and HR 3.72, HR95% CI 1.33 ~ 10.41, P = 0.012).

Multivariate cox regression after including conventional factors

To exclude the effect of traditional accepted risk factors, we put these factors with significant variables in log-rank test into multi-COX. As shown in Tables 3– 4, both M (HR 0.23, HR95%CI 0.08 ~ 0.66, P = 0.006 and HR 0.43, HR95%CI 0.20~0.91, P=0.026, respectively) and Stage IV (HR 7.12, HR95%CI 1.48 ~ 34.22, P = 0.014 and HR 3.79, HR95%CI 1.37 ~ 10.44, P = 0.010, respectively)were independently associated with OS and EFS of WT.

Table 3 Multivariate Cox regression of OS in WT after including conventional factorsTable 4 Multivariate Cox regression of EFS in WT after including conventional factorsConstruction and verification of the OS nomogram

To develop a survival nomogram, the total dataset was randomly divided into training and validation cohorts using a 7:3 ratio, as shown in Table S4. It can be observed that both datasets have baseline comparability. According to the outcomes of Cox analysis results (Table S5), 3 characteristics (NLR, stage and M) were eventually incorporated into the training cohort survival nomogram development.

As illustrated in Fig. 3a, the survival nomogram intuitively predicted the 3-, 5- and 8-year OS rates of WT patients in the training cohort. The C-index of the training and validation cohorts was 0.843 (95%CI 0.765 ~ 0.920) and 0.701 (95%CI 0.439 ~ 0.963), respectively. Specifically, ROC analysis showed that the survival nomogram correctly predicted 3 (AUC = 0.801), 5 (AUC = 0.845) and 8 (AUC = 0.914) year survival for WT patients. (see Fig. 3b). The calibration curve of the survival nomogram is shown in Fig. 3c, which approached the diagonal, indicating good calibration of the survival nomogram in the training cohort [15]. The DCA of the training and the validation cohort indicated that the clinical value of the nomogram is excellent. (Fig. 3d).

Fig. 3

The nomogram of predicting OS in patients with WT and the validation of this model. a Survival nomogram for the prediction of 3-year, 5-year, and 8-year OS in WT patients. The patient score for each axis is marked, summing them to obtain a total score. Place the total score on the total points axis and draw a vertical line,so that the 3-, 5-, and 8-year OS rates for WT were determined. b Predictive performance of the survival nomogram for Wilms tumor patients assessed by ROC curves. The y-axis represents sensitivity and the x-axis represents specificity for predicting the OS of WT. AUC was the area under the ROC curve. The AUC value closer to 1 indicates higher accuracy of the model. c The calibration curve of 3-year, 5-year, and 8-year OS was predicted by the training, validation cohort of WT patients. the diagonal on the chart indicates equal predicted and actual OS rates. The closer the calibration curve is to the diagonal line, the better the predicted OS aligns with the actual OS rate. d DCA curves employed for model validation. The x-axis is the threshold probability, and the y-axis is the net benefit. The red line indicates that no patients have died, and the light green line indicates that all patients have died. M absolute monocyte count, OS overall survival, AUC the area under the receiver operating characteristic curve