Table 4.
Related studies and strategies of BCa recurrence risk prediction during the past 20 years.
| Study | Patient | Treatment | Follow-up/years | Predictionmodel | Findings | Conclusion |
|---|---|---|---|---|---|---|
| Sylvester et al., 2006 (118) | 2596 NMIBC patients from 7 EORTC trials | TURBT + Intravesical treatment (78.4% of the patients) | Median follow-up of 3.9 years and maximum follow-up of 14.8 years | Univariate and multivariate analyses | The EORTC risk table was derived based on the number and size of tumors, prior recurrence rate, T category, carcinoma in situ, and grade. | EORTC risk table is a useful tool for the urologist to discuss the different options with the patient to determine the most appropriate treatment and frequency of follow-up. |
| Fernandez et al., 2009 (8) | 1062 NMIBC patients from 4 CUETO trials | TURBT + BCG with 12 instillations | 5 years | Univariate and multivariate analyses | The CUETO risk table was developed using gender, age, grade, tumor status, multiplicity and associated Tis. | The recurrence risks calculated by the CUETO table were lower than those obtained with EROTC table. |
| Seo et al., 2010 (122) | 251 patients from single center | TURBT + full-doze maintenance BCG | 5 years and 9 months | EORTC | C-index: 0.62 | The recurrence rate and progression rate were almost similar to the EORTC risk tables. However, the recurrence rate was low in the intermediate-risk group. |
| Xylinas et al., 2013 (120) | 4784 patients from 8 centers | TURBT +51% cohort of immediate single postoperative chemotherapy + 11% cohort of BCG | 4 years and 9 months | EORTC, CUETO | C-index: 0.60, 0.52 | Both models exhibited poor discrimination. Specific biomarkers should be exploited for improving the performance. |
| Xu et al., 2013 (48) | 363 NMIBC patients from single center | TURBT +79% cohort of immediate single postoperative chemotherapy + 100% cohort of the entire course of intravesical chemotherapy | 3 years | EORTC, CUETO | C-Index: 0.71, 0.66 | The EORTC model showed more value in predicting recurrence and progression in patients with NMIBC. |
| Kohjimoto et al., 2014 (121) | 366 NMIBC patients from single center | TURBT + BCG | 5 years | EORTC, CUETO | C-index: 0.51, 0.58 | Although both exhibited poorly for recurrence prediction, CUETO was a little better. |
| Vedder et al., 2014 (35) | 1892 NMIBC patients from 18 centers | TURBT +13~22% cohort of the entire course of intravesical chemotherapy+17~30% cohort of BCG + 0.55~0.61% cohort of Re-TURBT | 10 years | EORTC, CUETO | C-index: 0.56-0.59, 0.64-0.72 |
The discriminatory ability for BCa recurrence was unsatisfactory. |
| Cambier et al., 2016 (10) | 1812 NMIBC patients from 2 EORTC trials | TURBT + 1~3 years of maintenance BCG | 7 years 5 months | Updated EORTC | C-index: 0.59. | NMIBC patients treated with1~3 years of maintenance BCG had a heterogeneous prognosis among the high-risk patients, and early cystoscopy should be considered. |
| Dalkilic et al., 2018 (119) | 400 NMIBC patients from single center | TURBT + BCG (45.3% of the patients) | 5 years | EORTC, CUETO | C-index: 0.777, 0.703 | EORTC risk table was better than the CUETO table for the recurrence prediction. |
| Kim et al., 2019 (35) | 970 NMIBC patients from single center | TURBT + BCG | 5 years | New model, EORTC | AUC: 0.65, 0.56 | The new model developed by using gross hamartia, previous or concomitant upper urinary tract urothelial carcinoma, stage, grade, number of tumors, intravesical treatment performed better than the EORTC risk table. |
| Yajima et al., 2019 (48) | 91 NMIBC patients from single center | TURBT | 5 years | Inchworm sign (tumor stalk) on the DWI and ADC images | The progression rate of inchworm-sign-negative cases was significantly higher than that of inchworm-sign-positive cases, whereas there was no significant difference in the recurrence rate between two groups. | The absence of an inchworm sign and histological grade 3 were independent risk factors for progression. |
| Xu et al., 2019 (61) | 71 patients including 36 NMIBC patients and 35 MIBC patients from single center |
TURBT for the NMIBC patients and RC for the MIBC patients | 2 years | Radiomics nomogram developed based on the radiomics features extracted from T2WI, DWI, ADC, and DCE MRI data, and the clinical risk factors | The proposed radiomics nomogram exhibited good performance both in the training cohort (AUC: 0.915) and the validation cohort (AUC: 0.838) for the prediction of the BCa recurrence during 2 years after operation. | The proposed radiomics-clinical nomogram has potential in the preoperative prediction |