Vaibhav Sahai MBBS, Presenter: Nothing to Disclose

Ritu Randhawa Gill MBBS, Abstract Co-Author: Nothing to Disclose

Matthew A. Barish MD, Abstract Co-Author: Research grant, E-Z-EM, Inc, Lake Success, NY Consultant, Barco nv, Duluth, GA

To compare the efficacy of Computer-Aided Detection (CAD) volumetric measurements with unidimensional measurements in the evaluation of lung tumor growth

Patients with primary and metastatic lung lesions who underwent serial CT examinations at least 4 weeks apart were identified from the database of Tumor Imaging Metrics Core at Dana-Farber/ Harvard Cancer Center (DF/HCC). A total of 95 lesions in 32 patients from 12 DF/HCC Phase I/II trials with up to 5 lesions per patient were retrospectively analyzed. Maximum diameter and volume of the tumor were measured using FDA approved workstations (Voxar3D (Barco) and CAD Syngo LungCare (Siemens Medical Solutions) respectively). Patient response between time-points was assessed according to RECIST and Volumetric criteria (Complete Response (CR), Partial Response (PR), Stable Disease (SD), Progressive Disease (PD)). Lesion size, location and density were also evaluated to determine if they affected agreement between the two criteria.

Mean age was 58.75+13.6 years; 41% were female. As per RECIST, a threshold diameter of 10 mm with electronic calipers was used. 10 lesions that could not be assessed by CAD included those greater than 50 mm and/or with close approximation to the pleura. Growth assessment (PR as 30% or 65% decrease; PD as 20% or 44% increase according to RECIST and Volumetric criteria respectively) using unidimensional measurements disagreed with CAD volumetrics in 27% cases. Kappa statistic for response concordance was 0.39 (95% CI 0.11 to 0.67).

Although CAD volumetric measurement allows more objective evaluation with reduced inter-observer variability, growth assessment of lung tumors frequently disagrees with established unidimensional measurements.

CAD increases sensitivity and efficiency by aiding nodule detection, however results are inconsistent with RECIST criteria for treatment response evaluation in clinical trials.

Sahai, V, Gill, R, Barish, M, Comparison of Unidimensional and CAD Volumetric Measurements in Assessment of Lung Tumors.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5013631.html