Revisiting the role of the pathological grading in pediatric adrenal cortical tumors: results from a national cohort study with pathological review.

The prognosis of malignant pediatric adrenocortical tumors is closely related to disease stage, which is used to guide perioperative treatment recommendations. However, current scoring systems are inadequate to distinguish between benign and malignant adrenocortical tumors. Robust microscopic prognostic features that could help determine perioperative therapy are also lacking. The aim of this national study was to review the prognostic value of the Wieneke scoring criteria and Ki67 labeling index in unselected pediatric adrenocortical tumors. Using strict definitions previously defined by expert pathologists, a Wieneke score was re-attributed to each tumor after an independent and centralized review. In addition, Ki67 proliferation index was performed and reviewed for each case. A total of 95 cases were selected; all were treated between 2000 and 2018 and had histopathologic material and sufficient outcome-related information available. Localized disease was found in 88% of patients. Among those with advanced disease, 6% had tumor extension into adjacent organs and 5% had metastases at diagnosis. Median follow-up was 5 years and 3 months. The 5-year PFS was 82%, 95% CI [73%-91%]. Tumor stage significantly correlated with PFS (p < 0.0001). Tumor weight up to 200 g, extra-adrenal extension and initial non-complete surgical resection were statistically associated with worse outcomes. No recurrences nor metastases occurred when the Ki67 index was < 15%. Up to two of the following five factors including tumor necrosis, adrenal capsular invasion, venous invasion, mitotic count > 15/20 high-power fields, and Ki67 index > 15%, significantly correlated with worse outcomes. We propose a pathological scoring system incorporating the Ki67 index as part of a two-step approach after disease staging to guide adjuvant treatment in pediatric adrenocortical tumors, especially after incomplete resection. These results should be validated in an independent cohort.