基于影像组学的嗜铬细胞瘤/副神经节瘤转移风险综合预测模型研究

Pheochromocytoma/paraganglimoas (PPGLs) are neuroendocrine tumors arising from adrenal medullary and extra-adrenal chromaffin cells respectively, which posing a severe threat to human health. Now all PPGLs are considered to have metastasis potential. Patients would have a very poor prognosis once the tumors metastasize. However, effective ways of predicting PPGLs metastasis risk are still lacking. Radiomics has a widespread application in oncologic practice like tumor diagnosis, subtype classification, metastasis prediction and treatment response assessment due to its capability in quantitatively evaluating tumor heterogeneity. Our preliminary results showed that radiomics has a promising result in discriminating subclinical pheochromocytoma from lipid poor adrenal adenoma, reinforcing the feasibility of radiomics in PPGLs. Based on our pilot research, we will analyze from the aspects of PPGLs biomarkers expression, traditional imaging features, clinical characteristics and radiomics, aiming at constructing a radiomics-based PPGLs metastasis prediction model. Our findings will not only provide a solution to the challenging problem of PPGLs metastasis prediction but also provide new insights into the disease.

嗜铬细胞瘤/副神经节瘤(Pheochromocytoma/paraganglimoas,PPGLs)是一类分别起源于肾上腺髓质和肾上腺外嗜铬细胞的神经内分泌肿瘤，严重威胁人类健康。目前认为所有PPGLs均具有转移潜能。PPGLs一旦发生转移，患者预后极差，但目前仍缺乏有效预测其转移风险的方法。影像组学可以定量化评估肿瘤异质性，进而在肿瘤诊断、分型、转移预测及治疗反应评估等方面有着广泛的应用，我们的前期研究利用影像组学鉴别亚临床型嗜铬细胞瘤与乏脂型腺瘤，取得了良好的效果，证实了影像组学在PPGLs领域的良好适用性。在此研究基础上，本项目将从PPGLs分子标记物表达、传统影像学特征、临床基本特征、影像组学等多个方面进行分析，建立以影像组学为基础的PPGLs转移风险预测模型。研究结果不仅将为PPGLs转移风险预测提供新的解决办法，同时还为PPGLs的防治提供新思路。

嗜铬细胞瘤/副神经节瘤(Pheochromocytoma/paraganglimoas, PPGLs)是一类分泌儿茶酚胺的神经内分泌肿瘤。目前研究认为所有PPGLs均具有转移潜能。PPGLs一旦发生转移，患者预后极差，严重危害健康，但是目前临床上仍然缺乏有效预测PPGLs转移的方法。本项目从临床、病理、基因、影像组学等多个方面探讨如何预测PPGLs转移。我们研究发现Apelin、SDHB、CHGB和ERBB-2等肿瘤标志物在转移性和非转移性PPGLs中表达存在显著差异。同时我们采用NGS技术，研究PPGLs的基因突变情况。根据PPGLs的基因突变，将PPGLs分为PHT型和非PHT型PPGLs。我们利用临床资料，成功建立了预测PHT型PPGLs这一类更易转移的PPGLs。基于PPGLs的基因突变特征，结合患者的临床、病理、影像学资料，成功建立了预测PPGLs转移风险的模型。最后，我们收集我院和外院PPGLs患者的术前CT影像，从中提取影像组学特征，并结合临床、病理、基因等资料，有望建立一个高效的PPGLs转移风险预测的综合模型。