Department of Oncologic Imaging, Institut Bergonié, Regional Comprehensive Cancer, F-33076 Bordeaux, France; Department of Imaging, Pellegrin University Hospital, F-33300 Bordeaux, France; University of Bordeaux, UMR CNRS 5251, INRIA Project team Models in Oncology (Monc), F-33400 Talence, France.
Department of Medical Oncology, Institut Bergonié, Regional Comprehensive Cancer, F-33076 Bordeaux, France.
Eur J Radiol. 2022 Oct;155:110472. doi: 10.1016/j.ejrad.2022.110472. Epub 2022 Aug 12.
To investigate which acquisition, post-processing, tumor, and patient characteristics contribute the most to the value of radiomics features (RFs) in lung adenocarcinoma in order to better understand and order the potential sources of bias in radiomics studies in a multivariate setting.
This single-center retrospective study included all consecutive patients with newly-diagnosed lung adenocarcinoma treated between December 2016 and September 2018 who had pre-treatment contrast-enhanced CT-scan showing ≥ 2 target lesions per response evaluation criteria in solid tumors (RECIST) v1.1. All measurable lesions were manually segmented; 49 RFs were extracted using LIFEx v7.0.0. Afterwards, we reverted the usual radiomics approach (i.e., predicting a clinical outcome base on multiple RFs). To do so, for each RF, random forests and linear regression algorithms were trained using cross-validation to predict the RF value depending on the following variables: patient, mutational status, phase of CT-scan acquisition, discretization (binsize), lesion location, lesion volume, and best response obtained during the first line of treatment (partial response per RECIST vs other). The most important contributors to the value of reproducible RFs (intra-class correlation coefficient > 0.80) according to the best random forests model (selected via R-squared) were ranked.
101 patients (median age: 62.3) were included, with a median of 5 target lesions per patient (range: 2-10) providing 466 segmented lesions. Twenty-nine RFs were reproducible. The most important predictors of the reproducible RFs values were, in order: tumor volume, binsize, tumor location, CT-scan phase, KRAS mutation, and treatment response (average importance: 61.7%, 57.4%, 8.1%, 3.3%, 3%, and 2.7%, respectively). The treatment response and KRAS and EGFR/ROS1/ALK mutational status remained independently correlated with the RF value for 64.3%, 32.1%, and 50% reproducible RFs, respectively.
Tumor volume, location, acquisition and post-processing parameters should systematically be incorporated in radiomics-based modeling; however, most reproducible RFs do have significant relationships with mutational status and treatment response.
研究在肺腺癌中，哪些获取、后处理、肿瘤和患者特征对放射组学特征（RFs）的价值贡献最大，以便在多变量环境中更好地理解和排序放射组学研究中的潜在偏倚来源。
本单中心回顾性研究纳入了所有 2016 年 12 月至 2018 年 9 月期间新诊断为肺腺癌的连续患者，这些患者在 RECIST v1.1 标准下，治疗前的对比增强 CT 扫描显示每个反应评估标准中至少有 2 个靶病变。所有可测量的病变均进行手动分割；使用 LIFEx v7.0.0 提取 49 个 RFs。之后，我们反转了常用的放射组学方法（即，基于多个 RFs 预测临床结局）。为此，对于每个 RF，使用交叉验证通过随机森林和线性回归算法训练来预测 RF 值，具体取决于以下变量：患者、突变状态、CT 扫描采集阶段、离散化（binsize）、病变位置、病变体积以及一线治疗中获得的最佳反应（根据 RECIST 部分缓解 vs 其他的最佳反应）。根据最佳随机森林模型（通过 R 平方选择），根据可重复性 RF （组内相关系数> 0.80）的价值，对最重要的贡献者进行排名。
共纳入 101 例患者（中位年龄：62.3 岁），每位患者中位有 5 个靶病变（范围：2-10），共提供了 466 个分割病变。29 个 RFs 是可重复性的。可重复性 RF 值的最重要预测因素依次为：肿瘤体积、binsize、肿瘤位置、CT 扫描阶段、KRAS 突变和治疗反应（平均重要性分别为 61.7%、57.4%、8.1%、3.3%、3%和 2.7%）。治疗反应和 KRAS 以及 EGFR/ROS1/ALK 突变状态与 64.3%、32.1%和 50%的可重复性 RFs 的 RF 值仍分别独立相关。
在基于放射组学的建模中，应系统地纳入肿瘤体积、位置、采集和后处理参数；然而，大多数可重复性 RFs 与突变状态和治疗反应仍具有显著关系。
