摘要
HomeRadiologyVol. 300, No. 3 PreviousNext CommunicationsFree AccessLetters to the EditorRadiomics to Predict Invasiveness of Lung Adenocarcinoma in Part-Solid NodulesLinyu Wu, Chen Gao, Maosheng Xu Linyu Wu, Chen Gao, Maosheng Xu Author AffiliationsDepartment of Radiology, The First Affiliated Hospital of Zhejiang Chinese Medical University, 54 Youdian Road, Hangzhou 310006, PR ChinaThe First School of Clinical Medicine of Zhejiang Chinese Medical University, Hangzhou, Chinae-mail: [email protected]Linyu WuChen GaoMaosheng Xu Published Online:Jul 13 2021https://doi.org/10.1148/radiol.2021204661MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In Editor:We carefully reviewed the research conducted by Dr Wu and colleagues in the November 2020 issue of Radiology (1). This research aimed to build prediction models to aid in the accurate diagnosis of the invasiveness of lung adenocarcinoma in part-solid nodules (PSNs). Their research focused on analyzing the radiomic features both individually, and in combination from gross tumor volume (GTV), solid, ground-glass, and perinodular regions. The results indicated that the radiomic model combining the ground-glass and solid regions yielded the highest area under the curve of 0.98 and the highest accuracy of 93% in the test set. Although the results achieved by the radiomic models were promising, we have some questions about the method and would appreciate additional information regarding the study.First, the solid region was identified by using a thresholding method within the GTV (greater than −50 HU) in the study. However, there was no consensus about the appropriate definitions of the solid component of subsolid nodules. In this study, the maximum CT values of the vessel and the subsolid nodule were compared to determine the solid component of subsolid nodules (2). In addition, the Fleischer Society (3) recommended defining the size of the solid component in PSNs in its largest dimension visible on transverse CT sections and on thin sections with a mediastinal window setting. However, these were not mentioned in the article.Furthermore, the perinodular region was defined by extending the GTV by 5 mm from the nodule boundary in three dimensions and subtracting the GTV. However, if a nodule is near or under the costal or interlobar pleura, the perinodular region would be beyond the lung or lung lobe. Were these regions beyond the lung or lung lobe removed in the study? In addition, were large vessels and the main bronchus excluded in the perinodular region? If not, there is definite potential for bias here, which may affect the reliability of the results.Disclosures of Conflicts of Interest: L.W. disclosed no relevant relationships. C.G. disclosed no relevant relationships. M.X. disclosed no relevant relationships.Supported by the Zhejiang Provincial Natural Science Foundation of China (grant LSY19H180003).References1. Wu G, Woodruff HC, Shen J, et al. Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study. Radiology 2020;297(2):451–458. Link, Google Scholar2. Tu W, Li Z, Wang Y, et al. The "solid" component within subsolid nodules: imaging definition, display, and correlation with invasiveness of lung adenocarcinoma, a comparison of CT histograms and subjective evaluation. Eur Radiol 2019;29(4):1703–1713. Crossref, Medline, Google Scholar3. Naidich DP, Bankier AA, MacMahon H, et al. Recommendations for the management of subsolid pulmonary nodules detected at CT: a statement from the Fleischner Society. Radiology 2013;266(1):304–317. Link, Google ScholarReferences1. Wu G, Woodruff HC, Shen J, et al. Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study. Radiology 2020;297(2):451–458. Link, Google Scholar2. Tu W, Li Z, Wang Y, et al. The "solid" component within subsolid nodules: imaging definition, display, and correlation with invasiveness of lung adenocarcinoma, a comparison of CT histograms and subjective evaluation. Eur Radiol 2019;29(4):1703–1713. Crossref, Medline, Google Scholar3. Naidich DP, Bankier AA, MacMahon H, et al. Recommendations for the management of subsolid pulmonary nodules detected at CT: a statement from the Fleischner Society. Radiology 2013;266(1):304–317. Link, Google Scholar4. Liang J, Xu XQ, Xu H, et al. Using the CT features to differentiate invasive pulmonary adenocarcinoma from pre-invasive lesion appearing as pure or mixed ground-glass nodules. Br J Radiol 2015;88(1053):20140811. Crossref, Medline, Google Scholar5. Yue X, Liu S, Liu S, et al. HRCT morphological characteristics distinguishing minimally invasive pulmonary adenocarcinoma from invasive pulmonary adenocarcinoma appearing as subsolid nodules with a diameter of ≤3 cm. Clin Radiol 2018;73(4):411.e7–411.e15. Crossref, Google ScholarReferences1. Wu G, Woodruff HC, Shen J, et al. Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study. Radiology 2020;297(2):451–458. Link, Google Scholar2. Tu W, Li Z, Wang Y, et al. The "solid" component within subsolid nodules: imaging definition, display, and correlation with invasiveness of lung adenocarcinoma, a comparison of CT histograms and subjective evaluation. Eur Radiol 2019;29(4):1703–1713. Crossref, Medline, Google Scholar3. Naidich DP, Bankier AA, MacMahon H, et al. Recommendations for the management of subsolid pulmonary nodules detected at CT: a statement from the Fleischner Society. Radiology 2013;266(1):304–317. Link, Google ScholarResponseGuangyao Wu*,†, Henry C. Woodruff*,‡, Jing Shen†, Jianlin Wu† , Philippe Lambin*,‡Guangyao Wu*,†, Henry C. Woodruff*,‡, Jing Shen†, Jianlin Wu† , Philippe Lambin*,‡Author AffiliationsDepartment of Precision Medicine, Maastricht, the NetherlandsDepartment of Radiology and Nuclear MedicineGROW School for Oncology and Developmental Biology, Maastricht University Medical Center, Department of Radiology, Affiliated Zhongshan Hospital of Dalian University, 6 Jiefang Street, Dalian 116001, Chinae-mail: [email protected]We thank the authors for their interest and commentary regarding the definition of the solid and perinodular regions for PSNs in our study (1). We admit that the description of these parts was insufficient and emphasize that this correspondence has the ambition to go beyond guidelines and consensus.Indeed, there was no international consensus about how to define the solid region of PSNs. A previous study found that the CT window setting of −50 window length/2 window width resulted in an area under the curve greater than other window settings and the definition of visibility in the mediastinum window and indicated that the use of this window setting was the optimum way to identify the solid region and to distinguish invasiveness for subsolid nodules (2). On the basis of this publication, we selected −50 HU as a three-dimensional threshold for the solid region. Moreover, we evaluated semantic features in the lung and mediastinum windows, and the size of the solid component within PSNs was measured in the mediastinum window as recommended by Fleischer Society (3).There was also no consensus about how to define the perinodular region. We addressed this point with experts, and we concluded that, because we are focusing on lung nodules, the perinodular region should include only the area within the lung lobe. Therefore, the regions beyond the lung or lung lobe were removed from our study. As for the vessels and bronchus, they were excluded in the GTV and were included in the perinodular region. Although this may have caused potential bias, it is the simplest approach (no need for additional segmentation that could also introduce an error) and reflects the real information around the nodules. As previously reported, some CT features outside the nodules such as increased amount of blood vessels, vessel convergence, and air bronchograms may help the identification of invasive pulmonary nodules (4,5). However, the radiomic features of the perinodular region in our study did not contribute to predicting invasiveness for PSNs relative to radiomic signature on the basis of solid and ground-glass regions. Disclosures of Conflicts of Interest: G.W. disclosed no relevant relationships. H.C.W. Activities related to the present article: disclosed a grant from the Dutch Cancer Society. Activities not related to the present article: disclosed minority shareowner in OncoRadiomics. Other relationships: disclosed no relevant relationships. J.S. disclosed no relevant relationships. J.W. disclosed no relevant relationships. P.L. Activities related to the present article: disclosed grants from ERC, European Program, PREDICT, CHAIMELEON, TRANSCAN, RAIL; disclosed consulting fee from OncoRadiomics; disclosed support for travel from OncoRadiomics, BHV, Merck, ptTheragnostic/DNAmito, Elekta; disclosed money to author from Benelux Health Ventures. Activities not related to the present article: disclosed board memberships from Oncoradiomics, Convert Pharmaceuticals; consultancy from Benelux Health Ventures, ptTheragnostic-DNAmito, Convert Pharmaceuticals; payment for lectures from Varian, ptTheragnostic and Convert Pharmaceuticasl; money to author from patents; money to author from royalties; stock/stock options from Convert Pharmaceuticals, the Medical Cloud Company, Oncoradiomics; travel/accommodations/meeting expenses paid from Health Innovation Ventures; money to author for patents and copyrights pending, issued, licensed; money to author for patents and copyright royalties; and money to author for licensee from Oncoradiomics. Other relationships: disclosed no relevant relationships.References1. Wu G, Woodruff HC, Shen J, et al. Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study. Radiology 2020;297(2):451–458. Link, Google Scholar2. Tu W, Li Z, Wang Y, et al. The "solid" component within subsolid nodules: imaging definition, display, and correlation with invasiveness of lung adenocarcinoma, a comparison of CT histograms and subjective evaluation. Eur Radiol 2019;29(4):1703–1713. Crossref, Medline, Google Scholar3. Naidich DP, Bankier AA, MacMahon H, et al. Recommendations for the management of subsolid pulmonary nodules detected at CT: a statement from the Fleischner Society. Radiology 2013;266(1):304–317. Link, Google Scholar4. Liang J, Xu XQ, Xu H, et al. Using the CT features to differentiate invasive pulmonary adenocarcinoma from pre-invasive lesion appearing as pure or mixed ground-glass nodules. Br J Radiol 2015;88(1053):20140811. Crossref, Medline, Google Scholar5. Yue X, Liu S, Liu S, et al. HRCT morphological characteristics distinguishing minimally invasive pulmonary adenocarcinoma from invasive pulmonary adenocarcinoma appearing as subsolid nodules with a diameter of ≤3 cm. Clin Radiol 2018;73(4):411.e7–411.e15. Crossref, Google ScholarArticle HistoryPublished online: July 13 2021Published in print: Sept 2021 FiguresReferencesRelatedDetailsRecommended Articles Revisions to the TNM Staging of Lung Cancer: Rationale, Significance, and Clinical ApplicationRadioGraphics2018Volume: 38Issue: 2pp. 374-391Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter StudyRadiology2020Volume: 297Issue: 2pp. 451-458Visceral Pleural Invasion in Pulmonary Adenocarcinoma: Differences in CT Patterns between Solid and Subsolid CancersRadiology: Cardiothoracic Imaging2019Volume: 1Issue: 3Management of Nodules Attached to the Costal Pleura at Low-Dose CT Screening for Lung CancerRadiology2020Volume: 297Issue: 3pp. 710-718CT Characteristics and Pathologic Basis of Solitary Cystic Lung CancerRadiology2019Volume: 291Issue: 2pp. 495-501See More RSNA Education Exhibits The IASLC Lung Cancer Staging The New TNM (Eighth Edition): What You Have to KnowDigital Posters2018Shades of Gray: Pitfalls and Problem Solving for Subsolid DensitiesDigital Posters2018TNM 8th Edition in the Staging of Lung Adenocarcinoma Spectrum: What the Radiologist Needs to KnowDigital Posters2018 RSNA Case Collection Pneumothorax Ex VacuoRSNA Case Collection2022Granulomatosis with polyangiitisRSNA Case Collection2020Diffuse idiopathic pulmonary neuroendocrine cell hyperplasia with carcinoid tumorRSNA Case Collection2020 Vol. 300, No. 3 Metrics Altmetric Score PDF download