Establishment of a new prognostic model for lung cancer using peritumoral texture analysis

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K16693
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 52040:Radiological sciences-related
• Research Institution: Niigata University
• Principal Investigator: Yamazaki Motohiko 新潟大学, 医歯学系, 講師 (30772032)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 肺癌 / テクスチャ解析 / Radiomics / 予後予測 / EGFR遺伝子変異 / 腫瘍周囲肺

Outline of Final Research Achievements

Texture analysis refers to quantitative evaluation of image texture, and extracting many features from radiological images for diagnosis using texture analysis is called radiomics. Although utility of intratumoral radiomics has already reported, that of peritumoral radiomics is not well studied.
This study revealed that peritumoral radiomics is significantly associated with overall survival and EGFR gene mutations in lung cancer. In addition, a combination of intratumoral and peritumoral radiomics is found to improve predictability of prognosis and EGFR gene mutations compared to intratumoral radiomics alone. Particularly, peritumoral radiomics within 3 mm from the tumor boundary was important for prediction and was associated with various pathological prognostic factors such as cancer histology, tumor infiltrating lymphocytes, and tumor spread through air spaces.

Free Research Field

胸部CT画像診断

Academic Significance and Societal Importance of the Research Achievements

本研究によって原発巣周囲肺のRadiomicsの臨床的有用性が明らかとなり、原発巣のRadiomicsと併せて評価する事で、これまでよりも正確な肺癌のリスク分類が実現できると考えられた。周囲肺のRadiomicsと病理像との対比検討はまだ乏しいが、本研究によって周囲肺のRadiomicsが病理像を推定する一助となる事も示唆された。