Endocrine Abstracts

Background: Neuroendocrine tumors (NETs) are slow-growing, so radiological assessment of tumor progression using standard evaluation is challenging. RECIST 1.1 is commonly used to assess the response to treatment in solid oncology clinical trials. This study aims to assess the impact of RECIST 1.1 in a standard of care setting and compare its use to routine radiology reports on physician decision making and patient satisfaction.

Methods: We identified 50 patients (pts) from the Stanford NET clinic. We retrospectively used Stanford’s Tumor Response Assessment Criteria (TRAC) and mintLesion™, FDA-approved software, to apply RECIST 1.1 criteria to CTs or MRIs performed as part of standard of care imaging. To assess the impact of TRAC compared to standard radiology reports on physician decision-making, 20 physicians reviewed 5 de-identified patient cases. Physicians received standard radiology reports and answered survey questions regarding disease response; subsequently, they received TRAC reports for the same cases and responded to the same questions. The survey evaluated their interpretation of both the standard radiology report and TRAC report, if TRAC report would prompt changes in treatment, and measured their confidence in their decisions, comparing rates and reasons for treatment change. To assess pt satisfaction and understanding of their standard radiology report compared to TRAC report, pts were asked to complete 2 surveys, first after they received their standard radiology report and second after they were given TRAC report reviewed by a provider.

Results: Physician responses were analyzed using Kendall’s W to assess agreement in survey responses after separately reviewing standard and TRAC reports. Across all cases we found that the overall agreement of responses to interpretation after reviewing the standard report (0.77) and TRAC report (0.81) were similar; although physicians agreed within each method, their responses differed between standard and TRAC reports. Agreement on treatment changes was higher after reviewing TRAC report (0.70) than standard report (0.55). Patient surveys were analyzed using Wilcoxon signed rank test. Pts reported significantly improved understanding of their diagnosis (P = 0.011), felt more comfortable (P < 0.001), and were very satisfied with the way their imaging was explained with TRAC vs standard radiology report (P < 0.001).

Conclusions: Our study underscores the variability in physician interpretation of standard radiology reports and impact on treatment decisions in NETs. The results highlight several positive aspects of patient experience with TRAC compared to standard approach and suggests the benefits of using standardized response assessment for imaging in routine clinical practice.

ABSTRACT ID28344