68Ga-DOTATATE PET/CT parameters predict response to peptide receptor radionuclide therapy in neuroendocrine tumours

Purpose: [Lu-177]DOTATATE prolongs progression free survival (PFS) in metastatic neuroendocrine tumours (NETs). However, objective response rate is low. This, coupled with long duration of therapy and expense suggest need for better selection. We aim to assess whether baseline [Ga-68]DOTATATE-PET/CT parameters, and whether response assessment by PET accurately predicts clinical outcome to [Lu-177]DOTATATE. Experimental design: Retrospective study of patients receiving [Lu-177]DOTATATE was conducted. Patients were followed 3-monthly until disease progression. Four [Ga-68]DOTATATE-PET parameters (single lesion SUVmax, tumour to spleen and liver SUV ratios, and SUVmax-av using up to five target lesions in multiple organ sites) were determined at baseline and follow-up. The association between these PET parameters either at baseline, or any changes following treatment, and PET response criteria (PERCIST and modified PERCIST) to predict PFS were determined. Patients were followed 3-monthly until disease progression. Response was determined using RECIST 1.1. Baseline SSTR2 expression was assessed and compared with PET parameters. Results: 55 patients with metastatic NETs were identified predominantly small bowel (N = 18) and pancreatic (N = 8) in origin. 16 were low grade, 15 intermediate and 3 high grade. Response to PRRT (N = 47): partial response (PR) 28%, stable disease (SD) 60% progressive disease (PD) 13%. Response to PRRT predicted PFS: PR 71.8 months (95%CI: not achieved), SD 29.1 months (95%CI: 15.2-43.1), and PD 9.7 months (95%CI: 0-21.02). Baseline, single lesion SUVmax predicted both response and PFS with SUV cut-off of 13.0 giving high sensitivity and specificity. Tumoural SUVmax correlated with SSTR2 expression, Spearman's rho - 0.69, p < 0.01. Conclusions: Baseline single lesion SUVmax and SUVmax-av predicts response to [Lu-177]DOTATATE. Objective response following PRRT defines a subset of patients with markedly improved PFSBaseline SUVmax 13.0 defines a threshold below which patients have poor response to PRRT and worse PFS. SUV threshold analysis should be taken forward into prospective studies. Crown Copyright (C) 2019 Published by Elsevier B.V. All rights reserved.