OC-064 Pet Scanning in Gastric Cancer: a Comparison Of Avid Versus Non Avid Cancers

Introduction Only some gastric cancers (GC) are detectable on PET imaging (PET avid). Few studies have assessed the molecular and biological differences between PET avid and non-avid GCs. Thus there may be prognostic differences between the two groups that have not been established. The primary aim was to identify and measure clinical differences of PET avid and non-avid primary GCs. The secondary aim was to determine whether characteristic molecular differences exist between PET avid and non-avid primary GCs. Methods Participants selected for this study were individuals (male and female, any age) with a diagnosis of GC that attended the Peter MacCallum Cancer Centre, Victoria, Australia, during the period of 1992–2002 who had agreed to partake in research. A total of 52 primary GC cases who received an initial PET scan were identified and included in this study. Results From the 52 cases identified 29 tumours were PET avid and 23 were non-avid. PET avid tumours were mostly intestinal type, 86%; T stage 3, 46%, with no cases of T4. Avid tended to be N1, 47%. Suggesting they may spread via lymph rather than locally. An AJCC stage of Ib was the most frequent, 35%, but overall more were stage IIIa+b, 49%. Avid tumours were located along the greater and lesser curve of the body in equal proportion: 25% each. Non-avid tumours were mostly diffuse type, 68%; T stage 3, 41%, with 9% of T4. Non-avid tumours tend to be N0, 46%. Overall, AJCC stage IIIa was the most frequent, 33%, with 9% at stage IV. Non-avid tumours were found more in the antrum, 40%. Signet ring carcinomas were found to be significantly more likely to be PET non-avid, p = 0.017. Non-avid tumours were significantly, p = 0.004, less differentiated than avid. Progression free survival was significantly less in the avid group, survival of 808 versus 1208 days, p = 0.04. However, the overall survival showed little difference. From these results it is difficult to determine true survival difference. Larger studies are needed to investigate this further. The overall genetic profiles of the avid compared to non-avid were not significantly different. Conclusion Our results suggest that PET avid tumours are diagnosed earlier or that they are less locally invasive, and that non-avid tumours are locally invasive. PET does appear to provide valuable information regaurding the histological sub-type of the tumour, its likely differentiation, lymph node involvement and metastasis (stage). Information available from PET on the genomics of the tumour is still unclear but there does seem to be some difference in expression levels of genes in avid and non-avid tumours. Further studies with larger numbers are needed. The use of PET for diagnosis, preoperative staging and management planning is still uncertain. Disclosure of Interest None Declared