Identification of predictors of metastatical potential in paragangliomas to develop a prognostic score (PSPGL) - Supplemental data

Abstract

INTRODUCTION: Paragangliomas (PGL) are rare tumors in adrenal (AdPGL) and extra-adrenal locations (exAdPGL). Metastasis are found in ~ 5%-35% of PGLs, synchronously or many years after initial diagnosis. There are no reliable predictors of metastatic disease. OBJECTIVE: To construct a prognostic score of prediction of metastatic potential in PGL. A secondary objective was to identify prognostic predictors in metastatic PGL (MPGL). METHODS: Retrospective analysis of clinical and laboratorial data of patients with PGL seen in Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, between 1967 – 2019, and tumors histological e immunohistochemical (IHC) evaluation. Patients were divided in two groups, MPGL (with metastasis) or non-metastatic PGL (NMPGL – without metastasis during follow up ≥ 96 months after surgery). Univariate and multivariate analysis were performed to identify characteristics associated to metastatic behavior. A risk score was constructed based on the multivariate analysis coefficients. To assess predictors of disease aggressiveness in MPGL, this group was subdivided in “malignant” metastatic PGL (mMPGL – patients that deceased ≤ 72 months after surgery) and “benign” malignant PGL (bMPGL – patients that survived > 72 months after surgery). Kaplan Meier curves were built to estimate disease specific survival (DSS). RESULTS Out of 263 patients with PGL (mean age 39.7 ± 16.7 years-old, 56.3% female, 82.5% AdPGL, 15.2% exAdPGL and 2.7% AdPGL + exAdPGL), 35 patients were selected for MPGL group and 110 for NMPGL group. In univariate analysis, presence of symptoms, pathogenic variants in SDHB gene, vanil mandelic acid and noradrenaline (NAU) urine concentrations, PGLexAd, tumor size and scores in Pheochromocytoma of the adrenal gland scaled score (PASS) and Grading system for adrenal phaeochromocytoma and paraganglioma (GAPP) scores were associated with malignant behavior whereas IHC for chromogranin B showed inverse correlation. In multivariate analysis, four features remained related to metastatic disease and composed the Prognostic Score of Paragangliomas (PSPGL): necrosis (33 points), > 3 mitosis/10 HPF (28 points), extension into adipose tissue (20 points) and exAdPGL (19 points). The receiver operator characteristic (ROC) curve showed an area under the curve (AUC) for PSPGL of 0.970 and the cut off of 24 points had 89.5% sensibility and 91.5% specificity to metastatic disease diagnostic. The risk of metastatic disease was high for PSPGL ≥ 47 (> 95%), intermediate for 28 ≤ PSPGL ≤ 39 (~30% to 80%), low for PSPGL 19-20 (~10%), and almost null in PPGL=0. In MPGL IHC for Ki-67 ≥ 3% was related to DSS. CONCLUSIONS: PSPGL is easy to obtain and demonstrated a good performance in predicting metastatic potential. Ki-67-IHC was a good marker of aggressiveness in MPGL. Both can be useful tools in individualizing post-surgical follow up in patients with PGL. Key words: Paraganglioma. Pheochromocytoma. Prognosis. Metastasis. Score. PASS. Histology. Survival.