In this study, a total of 19,774 T1BC patients were enrolled between 2004 and 2015. Prior to propensity score matching (PSM), 19,520 patients were categorized into the non-involved group, while 254 patients were in the PUI group. Following PSM, 244 patients from the non-involved group were selected. The baseline characteristics of the cohort, both before and after PSM, are summarized in Table 1. No significant differences were observed in terms of age, race, marital status, number of tumors, radiotherapy, and chemotherapy recode between the two groups (P > .050). However, the non-involved group had a higher percentage of high-grade tumors compared to the PUI group (86.1% vs 75.6%, P < 0.001). Additionally, patients in the PUI group had a slightly higher rate of tumor size > 6cm compared to the non-involved group (3.9% vs 3.5%). The distribution of surgical approaches differed significantly between the two groups (P < 0.001), with a greater tendency for radical cystectomy (RC) (4.3% vs 2.2%) and pelvic exenteration (5.9% vs 2.3%) in the PUI group. After 1:1 PSM, adjusting for age, race, marital status, grade, tumor size, surgery, radiotherapy, and chemotherapy, differences in age, grade, and number of tumors still remained between the two groups (P < 0.05).

The External validation cohort included a total of 159 bladder cancer patients. Among them, 152 cases belonged to the non-involved group, while 7 cases were classified under the PUI group. Compared to the non-involved group, the PUI group showed a significantly higher percentage of patients who experienced fatal outcomes (100% vs. 34.9%, P = 0.044). Additionally, the survival time in the PUI group was significantly shorter than that of the non-involved patients (5.29 vs. 54.69, P < 0.001) (Supplementary Table 2).

The median follow-up time was 50 months, and a total of 8412 (43.09%) and 143 (56.3%) deaths from all causes were recorded in the non-involved group and PUI group, respectively. Among these deaths, 3588 (18.38%) and 77 (30.31%) patients specifically died from bladder cancer in the non-involved and PUI groups, respectively. Kaplan–Meier analysis revealed that after propensity score matching (PSM), the PUI group had significantly lower overall survival (OS) and cancer-specific survival (CSS) probabilities compared to the non-involved group (Fig. 1; P = 0.023 for OS; P < 0.001 for CSS). Similar results were found before PSM (Fig. 2). Subgroup analyses based on whether patients received chemotherapy showed that the PUI group had worse OS and CSS outcomes than the non-involved group in patients without chemotherapy (Fig. 3). Among patients receiving chemotherapy, those with PUI had worse CSS compared to those without. Univariate and multivariate Cox regression analyses were conducted to identify prognostic factors for cancer-specific mortality in patients with T1BC. The results indicated that age, marital status, PUI involvement, tumor size, surgery, and radiation were associated with CSS. Additional details can be found in Supplementary Table 2.

Cancer-specific survival and overall survival of T1 bladder cancer patients in PUI group and Non-involved group after PSM

Cancer-specific survival and overall survival of T1 bladder cancer patients in PUI group and Non-involved group before PSM

subgroup analyses of T1 bladder cancer patients after PSM based on whether patients received chemotherapy to compare survival outcome between PUI group and Non-involved group: A without chemotherapy; overall survival, B without chemotherapy; cancer-specific survival, C receiving chemotherapy; overall survival, D receiving chemotherapy; cancer-specific survival

After adjusting for age, race, marital status, grade, tumor size, and number of tumors in model I, PUI was confirmed as an independent risk factor for all-cause mortality (HR: 1.446, 95% CI: 1.210–1.726, P < 0.001) and cancer-specific mortality (HR: 1.816, 95% CI: 1.423–2.318, P < 0.001) in the multivariate analyses (Table 2). Similar results were obtained in model II, which additionally accounted for surgery, radiotherapy, and chemotherapy (HR: 1.470, 95% CI: 1.231–1.756, P < 0.001 for all-cause mortality; HR: 1.817, 95% CI: 1.423–2.320, P < 0.001 for cancer-specific mortality). After 1:1 PSM, multivariate Cox regression analyses showed a significant statistical difference in cancer-specific mortality (HR: 1.892, 95% CI: 11.255–2.852, P = 0.002 for model I; HR: 2.114, 95% CI: 1.382–3.233, P = 0.001 for model II) and all-cause mortality (HR: 1.520, 95% CI: 1.156–2.000, P = 0.003 for model I; HR: 1.618, 95% CI: 1.224–2.141, P = 0.001 for model II) between the PUI and non-involved groups.

A predictive model was used to estimate the 3- and 5-year cancer-specific survival (CSS) rates for patients with T1 bladder cancer. The results were visually presented as a nomogram and further validated in a separate group of patients (Fig. 4). The nomogram incorporated six risk factors known to be associated with CSS, namely age, surgery, radiotherapy, tumor size, primary tumor site, and marital status. Age was found to have the highest impact on CSS prognosis.

Nomogram predicting 3- and 5-year bladder cancer-specific survival probability for T1 bladder cancer patients. Variables include age, histology, surgery, radiotherapy, tumour size, PUI, and marital status. use: locate patient values at each axis. Draw a vertical line to the ‘‘Point’’ axis to determine how many points are attributed for each variable value. Sum the points for all variables. Locate the sum on the ‘‘Total Points’’ line. Draw a vertical line towards the 3Yrs.Surv. Prob. and 5Yrs.Surv. Prob, Prob. axes to determine respectively the 3-, and 5-year survival probabilities

The nomogram demonstrated a relatively good predictive ability for CSS, with a C-index of 0.715 (0.711–0.719) in the development cohort and 0.672 (0.667–0.677) in the validation cohort. The calibration curves (Fig. 5) showed a consistent match between actual observations and predicted outcomes for the probability of 3- and 5-year CSS, indicating good performance and reliability of the model.

The development cohort A Calibration plots of the nomogram for 3-year; B Calibration plots of the nomogram for 5-year; The validation group C Calibration plots of the nomogram for 3-year; D Calibration plots of the nomogram for 5-year