https://biometry.nci.nih.gov/cdas/approved-projects/1414/

Principal Investigator
Name
John Billings
Degrees
M.D.
Institution
Independent Researcher
Position Title
Investigator
Email
johnmbillings@gmail.com
Project Information
Study
NLST
Project ID
NLST-263
Title
RETROSPECTIVE ANALYSIS OF PULMONARY NODULES USING A NEW ALGORITHM FOR DETERMINING INTERVAL GROWTH, WITH SUBSEQUENT PROSPECTIVE ANALYSIS FOR BLINDED VALIDATION
Summary
Currently, small pulmonary nodules are measured in one or two transaxial dimensions for comparison with prior chest CT studies. While this method is currently used nationwide by general and thoracic radiologists, it is known that this method is suboptimal for detecting changes that may be clinically important, particularly in small nodules (< 5 mm). This in turn requires repeated CT scans and often invasive biopsies in patients whose pulmonary nodules are ultimately found to be benign. We propose to investigate alternative methods of analysis. Any new methodology needs to be tested, first retrospectively against a number of known lesions (both stable, and changing). Once validated retrospectively, then a prospective analysis would then be performed. It is proposed that improving the method of analysis offers a number of benefits. One benefit is earlier detection of a clinically meaningful change. Another benefit is reduction of false positives for lesions that are actually stable, which translates to less unneeded follow-up imaging, and less anxiety for the patient. Reducing unneeded imaging directly relates to health care cost savings. Earlier detection of a changing lesion could feasibly reduce overall healthcare costs relating to treatment, this, however, would need to be demonstrated in a different study.

In this pilot study we propose to analyze a set of patients who have undergone screening for and evaluation of pulmonary nodules identified through low dose lung cancer screening CT scans. By evaluating serial CT scans we will test methodologies to distinguish benign pulmonary nodules from those in need of further evaluation and workup. Sets of CT scans with patient identifiers removed will be provided for review and interpretations based on our methodologies by one reviewer and will be compared to the known outcomes of the patients.

Aims
Primary Objective - To demonstrate the comparative sensitivity and specificity (versus current methods) of algorithms to evaluate lung lesion progression with examinations prospectively.

Secondary Objective - To demonstrate the results of one or more algorithms on a retrospective data set for validation.

Collaborators
Stephan Grant, MD, JD
Josh Tan, MS