Prognostic significance of histopathological response to preoperative chemotherapy in unilateral Wilms' tumor: An analysis of 899 patients treated on the SIOP WT 2001 protocol in the UK‐CCLG and GPOH studies

In the SIOP Wilms' tumor (WT) studies, preoperative chemotherapy is used as primary treatment, and tumors are classified thereafter by pathologists. Completely necrotic WTs (CN‐WTs) are classified as low‐risk tumors. The aim of the study was to evaluate whether a subset of regressive type WTs (RT‐WTs) (67%‐99% chemotherapy‐induced changes [CIC]) showing an exceptionally good response to preoperative chemotherapy had comparably excellent survivals as CN‐WTs, and to establish a cut‐off point of CIC that could define this subset. The study included 2117 patients with unilateral, nonanaplastic WTs from the UK‐CCLG and GPOH‐WT studies (2001‐2020) treated according to the SIOP‐WT‐2001 protocol. There were 126 patients with CN‐WTs and 773 with RT‐WTs, stages I‐IV. RT‐WTs were subdivided into subtotally necrotic WTs (>95% CIC) (STN‐WT96‐99) (124 patients) and the remaining of RT‐WT (RR‐WT67‐95) (649 patients). The 5‐year event‐free survival (EFS) and overall survival (OS) for CN‐WTs were 95.3% (±2.1% SE) and 97.3% (±1.5% SE), and for RT‐WTs 85.7% (±1.14% SE, P < .01) and 95.2% (±0.01% SE, P = .59), respectively. CN‐WT and STN‐WT96‐99 groups showed significantly better EFS than RR‐WT67‐95 (P = .003 and P = .02, respectively), which remained significantly superior when adjusted for age, local stage and metastasis at diagnosis, in multivariate analysis, whereas OS were superimposable (97.3 ± 1.5% SE for CN‐WT; 97.8 ± 1.5% SE for STN‐WT96‐99; 94.7 ± 1.0% SE for RR‐WT67‐95). Patients with STN‐WT96‐99 share the same excellent EFS and OS as patients with CN‐WTs, and although this was achieved by more treatment for patients with STN‐WT96‐99 than for patients with CN‐WT, reduction in postoperative treatment of these patients may be justified.


| INTRODUCTION
The outcomes for patients with Wilms' tumors (WTs) have significantly improved over the last decades, with >90% overall survival for those with localized, and 80% for those with metastatic nonanaplastic WT. [1][2][3] It is now increasingly important to refine the risk groups and find prognostic factors which identify WT subgroups requiring more aggressive treatment, as well as those who need less treatment to reduce the long-term sequelae and improve patients' quality of life. In the Children's Oncology Group (COG) trials and studies, a selected group of patients with stage I WTs which are regarded as very lowrisk WTs are treated with surgery only. [4][5] In the International Society of Paediatric Oncology (SIOP) Nephroblastoma Trials and Studies, preoperative chemotherapy has been used in the treatment of WTs and responsiveness to preoperative chemotherapy has been considered for tumor risk and treatment stratification. The SIOP 9 study has demonstrated that completely necrotic WTs (CN-WTs) had a significantly better prognosis than other subtypes 6 and they have been moved to the low-risk group in the subsequent SIOP classifications. 7,8 The regressive type WT (RT-WT), defined as WTs showing 67%-99% of chemotherapy-induced changes (CIC), has been placed in the intermediate-risk group. 8 Thus, an important stratification and treatment boundary depends on the absence or presence of any viable tumor at all. However, no study has ever scrutinized whether the presence of a small amount of viable tumor is associated with good outcomes comparable to those of CN-WT. In contrast, in bone tumors the histologic response to neoadjuvant chemotherapy in terms of the extent of necrosis has been established as a prognostic indicator for many years. [9][10][11][12] Recently, a similar approach has been suggested in soft tissue sarcomas (STS), 13 although the results of different studies were difficult to compare since there is no standardized scheme for histopathologic assessment of tumor response for STS, and no optimum cut-off to differentiate responders from nonresponders. Further, it is unclear whether the cut-off of prognostic significance is similar in different histological subtypes of STS, anatomic primary sites and treatment modalities (radiotherapy, chemotherapy, chemotherapy schedules). Some studies demonstrated favorable outcome using a cut-off ≤5% of viable tumor cells, [14][15][16] but others found no correlation between the extent of necrosis and clinical outcome. 17,18 The multiple assessment limitations of STS do not represent such a challenge in WT, making it an ideal candidate for the assessment of the correlation between histopathologic response to preoperative treatment and prognosis. Preoperative chemotherapy given in the SIOP studies is standardized, as is the sampling of tumor, and the assessment performed to a benchmarked standard by a small group of experts, through a system of central pathology review. 19,20 The aim of our study was to evaluate whether patients with RT-WTs showing a particularly good response to preoperative chemotherapy had comparably excellent survivals as seen in CN-WTs and could be candidates for reduced treatment.

| Study population
The cases were identified from the UK Children's Cancer and Leukaemia For different results and analyses, only cases with relevant information available were included.

| Histologic assessment
A retrospective analysis of WTs was done to identify cases that were either CN-WT (ie, tumors that showed 100% CIC) or RT-WTs (tumors with 67%-99% CIC). In order to be able to assess whether there were differences in survival within the RT-WT group, we further subdivided them into subtotally necrotic WTs (STN-WT96-99) (defined as WTs showing >95% of CIC) and the remaining of the RT-WTs (RR-WT67-95) (tumors showing 67%-95% of CIC). Finally, the RR-WT67-95 group was subdivided into RR-WT67-89 and RR-WT90-95% groups which were then analyzed separately.

All cases were sampled according to the SIOP-WT-2001 Study
Pathology protocol and submitted for central pathology review for diagnosis, risk classification and abdominal tumor staging, 8 performed by the SIOP-UK (GMV) and SIOP-GPOH (CV) Pathology Panels. The sampling of lymph nodes was recorded as "yes" or "no/unknown." The number of slides submitted for central pathology review was readily available in 1203 cases. It varied from 9 to 94 (median 29).

| Treatment
All patients were treated according to the SIOP-WT-2001 Study protocol (Table S1).
Follow-up information was obtained from the Study databases containing information documented in case report forms specific to each phase of diagnosis, treatment and follow-up and received regularly from the participating centers.

| Statistical analysis
Statistical analysis was performed using SPSS statistical software

| Patient outcomes
The median follow-up time was 5.8 years (mean 6.3 years, range from 9 to 178 months). The 5-year EFS and OS estimates for all analyzed groups are presented in Table 2 and Figures 1 and 2.
There was no significant difference in OS of patients from all groups who relapsed with localized or metastatic WTs (P = .3).
Six patients without relapse died: two due surgery-related postoperative complications, one patient developed glioblastoma, one due to acute myeloid leukemia, one died during stem cell transplantation, and one patient died 32 months after the diagnosis, recorded only as "tumor-related death, with no relapse."

| Patterns of recurrence
The types of relapses in all groups and stages are presented in Table 4.
Distant relapses were more common than local relapses (P < .00001). RR-WT90-95 groups, but there were no differences between them, so they were not further analyzed separately (Table S2).
There was no difference in relapses between the CN-WT group and the No relapse 44 100 --12 100 53 91 The present study readdressed the question of whether there were patients within the RT-WTs group who showed outcomes comparable to CN-WTs, so they could be candidates for treatment reduction.
The prevalence of CN-WT in our study was 6.0%, which was significantly lower than in the SIOP 9 study (10%) (P = .0002). 6 But, since the SIOP 9 study, subsequent studies have shown that only 4%-6% of WT were completely necrotic type. [22][23][24] Although in the SIOP The prognosis for patients with CN-WTs in the SIOP 9 study was excellent, with OS of 97% for patients with localized and 100% for patients with metastatic WT, 6