68Ga-DOTATATE PET/CT to predict response to peptide receptor radionuclide therapy (PRRT) in neuroendocrine tumours (NETs).

4093 Background: PRRT represents a step change in NET management, significantly improving survival. However, objective response to PRRT, approximately 20%, is poor. There are no predictive biomarkers of response. Uptake on 68Ga-DOTATATE PET/CT imaging is used to assess patient suitability for PRRT, highlighting the presence of somatostatin receptors (SSTR) to which PRRT selectively binds. We hypothesise that the density of SSTRs, as defined by a minimum SUV uptake, predicts for response to PRRT. Methods: 54 patients underwent PRRT. Modified PERCIST assessment was performed: up to 2 target lesions per organ were identified and volume of interest drawn. Maximum 5 targets were counted. Average SUV (SUVave) was calculated by dividing sum of SUVmax of target lesions by number of lesions. Response was determined by RECIST 1.1. Ki67 and SSTR2 expression were assessed on tumour samples and compared with SUVave. Results: Response to PRRT: partial response (PR) 26%, stable disease (SD) 40% progressive disease (PD) 12%. Response to PRRT predicted progression free survival (PFS) with patients experiencing PR having a PFS 2.5x that of those with SD, and almost 20x as long as PD. Using ROC curve analysis, SUVave of 21.6 predicted for tumour response with high sensitivity (0.74) and specificity (1.0), p = 0.15, 95% CI 0.71-3.96. No association between baseline SUVave and SSTR2 or Ki-67 was observed. SUVave > 21.6 was an independent predictor of clinical outcome. Conclusions: Objective response to PRRT defines a subset of patients with markedly improved PFS. SUVave 21.6 defines a threshold below which patients have a poor response to PRRT. This threshold should be taken forward into prospective study.