Little is known about the factors that enable the mobilisation of human mesenchymal stem cells (MSC) from the bone marrow into the blood stream and their recruitment to and retention in the tumour. We found specific migration of MSC towards growth factors present in pancreatic tumours, such as PDGF, EGF, VEGF and specific inhibitors Glivec, Erbitux and Avastin interfered with migration. Within a few hours, MSC migrated into spheroids consisting of pancreatic cancer cells, fibroblasts and endothelial cells as measured by time-lapse microscopy. Supernatant from subconfluent MSC increased sprouting of HUVEC due to VEGF production by MSC itself as demonstrated by RT-PCR and ELISA. Only few MSCs were differentiated into endothelial cells in vitro, whereas in vivo differentiation was not observed. Lentiviral GFP-marked MSCs, injected in nude mice xenografted with orthotopic pancreatic tumours, preferentially migrated into the tumours as observed by FACS analysis of green fluorescent cells. By immunofluorescence and intravital microscopic studies, we found the interaction of MSC with the endothelium of blood vessels. Mesenchymal stem cells supported tumour angiogenesis in vivo, that is CD31 þ vessel density was increased after the transfer of MSC compared with siVEGF-MSC. Our data demonstrate the migration of MSC toward tumour vessels and suggest a supportive role in angiogenesis.

The data provide new insights into resistance mechanisms of TICs and suggest the combination of sulforaphane with TRAIL as a promising strategy for targeting of pancreatic TICs.

PURPOSE The anti–B-cell maturation antigen BiTE molecule AMG 420 was assessed in patients with relapsed/refractory multiple myeloma. PATIENTS AND METHODS In this first-in-human study, up to 10 cycles of AMG 420 were given (4-week infusions/6-week cycles). Patients had progression after ≥ 2 lines of prior therapy and no extramedullary disease. Minimal residual disease (MRD) response was defined as < 1 tumor cell/104 bone marrow cells by flow cytometry. RESULTS Forty-two patients received AMG 420 at 0.2-800 μg/d. Median age was 65 years, and median disease duration was 5.2 years. Median exposure was 1 cycle (range, 1-10 cycles) and 7 cycles (range, 1-10 cycles) for responders. Patients discontinued for disease progression (n = 25), adverse events (AEs; n = 7), death (n = 4), completion of 10 cycles (n = 3), and consent withdrawal (n = 1). Two patients remain on treatment. There were 2 nontreatment-related deaths from AEs, influenza/aspergillosis and adenovirus-related hepatitis. Serious AEs (n = 20; 48%) included infections (n = 14) and polyneuropathy (n = 2); treatment-related serious AEs included 2 grade 3 polyneuropathies and 1 grade 3 edema. There were no grade ≥ 3 CNS toxicities or anti-AMG 420 antibodies. In this study, 800 μg/d was considered to not be tolerable because of 1 instance each of grade 3 cytokine release syndrome and grade 3 polyneuropathy, both of which resolved. The overall response rate was 31% (n = 13 of 42). At the maximum tolerated dose (MTD) of 400 μg/d, the response rate was 70% (n = 7 of 10). Of these, five patients experienced MRD-negative complete responses, and 1 had a partial response, and 1 had a very good partial response; all 7 patients responded during the first cycle, and some responses lasted > 1 year. CONCLUSION In this study of AMG 420 in patients with relapsed/refractory multiple myeloma, the response rate was 70%, including 50% MRD-negative complete responses, at 400 μg/d, the MTD for this study.