ESTRO 2024 - Abstract Book

S2226

Clinical - Upper GI

ESTRO 2024

IOV was worse for case 1 than case 2. For Case 1, GTVp conformity improved with addition of MRI (JCI= 0.66 with CT vs 0.71 with CT+MRI) but worsened for Case 2 (JCI = 0.88 for CT vs 0.83 with CT+MRI). Conformity of CTVstomach was high for CT, with MRI adding little. Regarding qualitative feedback, for Case 1, addition of MRI improved ease of TVD of GTVp in 72.7% of observers, with a mean ‘ease of delineation’ rating (rated 1 -10, 1= difficult, 10=easy) of 4.5 for CT vs 6.7 for CT+MRI. For CTVstomach, mean rating was 7.2 for CT vs 7.9 for CT+MRI. 27.3% observers reported no change in ease rating. For Case 2, MRI improved ease of TVD of GTVp in 90% of observers, with mean ease rating of 5.7 for CT vs 7.8 for CT+MRI. For CTVstomach rating was 5.7 for CT vs 6.5 with CT+MRI. 10% observers reported no change in ease rating. Common perceived areas of difficulty using CT alone were differentiation of tumour from normal stomach wall (n=6) and identification of pylorus (n=8), involved lymph nodes (n=7) and extra-gastric extension (n=3). N=15 (71.4%) reported that MRI improved visualisation for TVD, with DWI highlighted as useful. In contrast, n=6 observers found correlation of MRI to CT difficult given different stomach-filling and lack of co-registration. Potential confounders include variation in prior gastric TVD experience among observers, and differences in tumour location, volume and extra-gastric extension between cases. Limitations include small sample size, variable observer experience in MRI interpretation, and variation in stomach position between CT and MRI. None rated MRI as inferior to CT alone for GTVp TVD.

Conclusion:

In this first study to assess IOV in gastric GTVp delineation, we have shown significant variation exists using CT alone. Addition of MRI subjectively improved ease of delineation, but impact on IOV was variable, possibly related to prior gastric TVD experience and/or interpretation of MRI. In addition to addressing these issues, further IOV studies should utilise prospective planning CT with co-registered MRI in treatment position, with appropriate patient preparation, to optimally evaluate the impact of MRI in this setting.