P4462 - Correlation of Tumoral and Peri-Tumoral Findings Observed on CT Scan and Endoscopic Ultrasonography in Patients With Gastrointestinal Stromal Tumors

Nitish Mittal, MD¹, Faisal Ali, MD², Ingrid Schwartz, MD³, Iyad Al-bustami, MD, MPH(c)⁴, Jennifer Ma, MD¹, Stephen Glombicki, MD², Kazi Haque, MD⁵, William Ross, MD³, Brian Weston, MD³, Emmanuel Coronel, MD³, Phillip Ge, MD³, Jeffrey H.. Lee, MD, MPH, FACG^3
1University of Texas Health, McGovern Medical School, Houston, TX; ²University of Texas, Houston, TX; ³MD Anderson Cancer Center, Houston, TX; ⁴Brooklyn Hospital Center, Houston, TX; ⁵McGovern Medical School at UTHealth, Pearland, TX

Introduction: Gastrointestinal stromal tumors (GISTs) are often found on cross-sectional imaging which is followed by endoscopic ultrasonography (EUS). Correlation of tumoral and peritumoral findings noted on CT and EUS, and their prognostic value for patients with GISTs has not been explored.

Methods: We conducted a retrospective study of patients diagnosed with GIST from 2003 to 2023 at a tertiary cancer center. Association of imaging features was explored with tumor recurrence or progression using Chi Square test. Pearson correlation was performed to assess the concordance in CT and EUS findings.

Results: A total of 100 were included in this study. Most patients were female (54%) and Caucasian (71%), mean age of 63.4±12.8. Most GISTs were in the stomach (79%); within the stomach, most GISTs were in the gastric body (56%). Duodenal and esophageal GISTs constituted 3% and 14% of cases, respectively. The median tumor size was 3.6 cm on CT and 3.5 cm on EUS. Regular tumor contour was noted in 34% and 33%, irregular tumor contour in 10% and 5.4%, and tumor ulceration in 12% and 10% of CT and EUS, respectively. Peri-tumoral lymphadenopathy was noted in 8% and 6.3% of CT and EUS, whereas vascular involvement of tumor (feeder or draining vessel) was noted on 7% and 5.4% of CT and EUS, respectively. Tumor ulceration was noted in a higher proportion of patients who experienced post-resection recurrence (21% vs 5%; p= 0.01), whereas regular lesion contour was noted in a higher proportion of patients who did not experience post-resection recurrence (3% vs 19%; p= 0.02). There was moderate correlation between CT and EUS-based tumor size measurement (p=0.005); tumor size measurement with EUS strongly correlated (R= 0.8; p< 0.001), whereas tumor size measurement with CT moderately correlated with resected tumor specimen diameter (R= 0.6; p< 0.001). The findings of peritumoral lymphadenopathy and tumor ulceration on EUS moderately correlated with the of CT findings (R= 0.4; p< 0.001). There was no correlation between tumor contours as seen on CT and EUS.

Discussion: CT and EUS demonstrated strong correlation with tumor dimensions measured on surgically resected specimens. CT and EUS also demonstrated moderate correlation for findings of peritumoral lymphadenopathy and tumor ulceration, but not for vascular involvement by GISTs or tumor contours. Future studies are needed to identify factors that can improve correlation between both non-invasive imaging modalities.



Disclosures:


Nitish Mittal, MD¹, Faisal Ali, MD², Ingrid Schwartz, MD³, Iyad Al-bustami, MD, MPH(c)⁴, Jennifer Ma, MD¹, Stephen Glombicki, MD², Kazi Haque, MD⁵, William Ross, MD³, Brian Weston, MD³, Emmanuel Coronel, MD³, Phillip Ge, MD³, Jeffrey H.. Lee, MD, MPH, FACG³. P4462 - Correlation of Tumoral and Peri-Tumoral Findings Observed on CT Scan and Endoscopic Ultrasonography in Patients With Gastrointestinal Stromal Tumors, ACG 2024 Annual Scientific Meeting Abstracts. Philadelphia, PA: American College of Gastroenterology.