Upper GI 27

Aims: Positron emission tomography is increasingly used in the detection, staging and assessment of response to therapy in patients with upper GI cancer, but no agreement exists on the method of quantification of scans to allow comparison of results. Quantification can be achieved by the calculation of simple ratio measurements, uptake values or more complex influx constants, and we aim to determine which method of quantification is required to obtain useful clinical information. Methods: A total of 40 patients underwent FDG-PET imaging. All patients underwent static imaging with the calculation of tumour/liver ratios (TLR) and standardized uptake values (SUV) for quantification. Nineteen patients had both static and more complex dynamic imaging, allowing in addition the calculation of K, the influx constant of FDG. A total of 16 patients had PET scans pre- and post-chemotherapy. Results: FDG-PET detected 39 of 40 lesions, not detecting an area of linitus plastica in one patient. Correlation coefficients were calculated between methods of quantification. A correlation coefficient of 0.92 existed between TLR and SUV. Agreement was poor between TLR and SUV compared with K-values (0.51 and 0.41). In contrast, the correlation coefficient was 0.9 between TLR and K-value quantification when detecting change in activity following chemotherapy. Changes in activity ranged from a 15 per cent increase to a complete response following chemotherapy. Conclusions: FDG-PET is reliable in the detection of upper GI tumours. It shows great potential in the noninvasive detection of response to chemotherapy, and this can be achieved by a shorter imaging time and less complex calculation of tumour/liver ratios.