BACKGROUND: The aim of this study was to analyze parameters to predict tumor invasiveness according to high-resolution computed tomography and positron emission tomography in patients with clinical stage IA lung adenocarcinoma. METHODS: A total of 122 patients with clinical stage IA lung adenocarcinoma were enrolled in the study. Receiver operating characteristic (ROC) curves were constructed for three factors--the degree of solid tumor component (solid%), maximum standard uptake value (SUVmax) and tumor size--and cutoff values were determined to reveal the highest sensitivity and specificity to diagnose tumor invasiveness. We created an algorithm for detecting tumor invasiveness (model 1). The data for the three factors were combined and their ROC curves constructed (model 2). A prospective study was conducted to validate the utility of these models. RESULTS: Multivariate analysis identified solid%, SUVmax, and tumor size as potentially important predictors of tumor invasiveness. In the ROC curve analysis, solid% (area under the curve was 0.882) had the largest area under the curve, followed by the SUV (0.867) and tumor size (0.747). The combination assay using all three factors had the highest sensitivity and specificity for prediction (0.902). Models 1 and 2 were applied to the prospectively enrolled cases, and their utility was reviewed. Both models showed 100% sensitivity, with model 2 showing a slightly higher diagnostic value than model 1. CONCLUSIONS: The solid portion ratio was a more powerful clinical predictor for lymphovascular invasion than the SUVmax. Our novel scoring model for tumor invasiveness can be employed for preoperative assessment of tumor invasiveness.
BACKGROUND: The aim of this study was to analyze parameters to predict tumor invasiveness according to high-resolution computed tomography and positron emission tomography in patients with clinical stage IA lung adenocarcinoma. METHODS: A total of 122 patients with clinical stage IA lung adenocarcinoma were enrolled in the study. Receiver operating characteristic (ROC) curves were constructed for three factors--the degree of solid tumor component (solid%), maximum standard uptake value (SUVmax) and tumor size--and cutoff values were determined to reveal the highest sensitivity and specificity to diagnose tumor invasiveness. We created an algorithm for detecting tumor invasiveness (model 1). The data for the three factors were combined and their ROC curves constructed (model 2). A prospective study was conducted to validate the utility of these models. RESULTS: Multivariate analysis identified solid%, SUVmax, and tumor size as potentially important predictors of tumor invasiveness. In the ROC curve analysis, solid% (area under the curve was 0.882) had the largest area under the curve, followed by the SUV (0.867) and tumor size (0.747). The combination assay using all three factors had the highest sensitivity and specificity for prediction (0.902). Models 1 and 2 were applied to the prospectively enrolled cases, and their utility was reviewed. Both models showed 100% sensitivity, with model 2 showing a slightly higher diagnostic value than model 1. CONCLUSIONS: The solid portion ratio was a more powerful clinical predictor for lymphovascular invasion than the SUVmax. Our novel scoring model for tumor invasiveness can be employed for preoperative assessment of tumor invasiveness.
Authors: K Takamochi; K Nagai; J Yoshida; K Suzuki; Y Ohde; M Nishimura; S Sasaki; Y Nishiwaki Journal: J Thorac Cardiovasc Surg Date: 2001-08 Impact factor: 5.209
Authors: M Noguchi; A Morikawa; M Kawasaki; Y Matsuno; T Yamada; S Hirohashi; H Kondo; Y Shimosato Journal: Cancer Date: 1995-06-15 Impact factor: 6.860