Background: Metastatic neuroendocrine neoplasms (mNEN) originating in the gastrointestinal tract are frequently slow growing yet both symptom and disease control remain important. Treatment options include resection, systemic somatostatin analogues (SSA), and systemic peptide receptor radionuclide therapy (PRRT). Additional options are needed; we have explored intralesional (IL) rose bengal disodium (PV-10), an investigational autolytic immunotherapy that can yield immunogenic cell death and disease-specific functional adaptive immunity. Methods: This phase 1 study evaluated safety, tolerability and impact on symptoms and biochemical markers resulting from IL PV-10 administered percutaneously to hepatic lesions in patients (pts) with progressive mNEN not amenable to resection or other potentially curative therapy. Eligible lesion(s) were 1.0 - 3.9 cm in longest diameter with amount of PV-10 administered proportional to size. Cohort 1 (n = 6 pts) received PV-10 to a single lesion per treatment cycle; Cohort 2 (n = 6) could receive injection to multiple lesions per treatment cycle. Pts could receive further PV-10 ≥6 weeks after prior injection. The primary endpoint was safety. Secondary endpoints included objective response rate (ORR) assessed by contrast enhanced CT (RECIST 1.1) and 68Ga-DOTATATE PET, biochemical response (CgA) and patient-reported outcome (EORTC QLQ-C30 and GI.NET21 QOL instruments). Results: Twelve pts were enrolled, 50% male, median age 66 yrs (range 47-79). Primary sites: 7 small bowel, 2 pancreas, 1 caecal, 2 unknown; grade: Gd1 = 5, Gd2 = 7. All pts had received SSA and PRRT as part of previous therapy and all had symptomatic, progressive disease. Median CgA was 1585 (range 35-10370). One lesion was injected per cycle for all 12 pts; none were suitable for multiple injections. One pt received 4 sequential PV-10 treatment cycles, 3 received 2 cycles, and 8 received 1 cycle. Toxicity was consistent with experience in other hepatic malignancies: post-procedure pain was reported by most pts; grade 3 photosensitivity reaction occurred in 1 pt; and grade 1 elevation of hepatic enzymes attributed to PV-10 occurred in 2 pts, resolving by day 7. Additionally, carcinoid flare occurred in 1 pt. ORR of injected lesions was 42%; patient-level disease control was 84%. Estimated PFS was 9.2 months; median OS was 22.5 months. CgA remained stable in 10 pts and upregulation of NK and activated CD4+ T lymphocytes was observed post-injection. QOL data at months 1 and 3 showed stable or improved carcinoid symptoms and global health status in 9 pts. Conclusions: PV-10 elicited no safety concerns with encouraging evidence of both local and systemic disease and symptom control in a heavily pre-treated population. Multiple cycles were delivered safely in suitable patients. Adaptive immune upregulation is consistent with other solid tumors and supports potential systemic benefit. Clinical trial information: NCT02693067